WeRelate talk:Source renaming project

404 Error [22 August 2009]

Dallan, I get a 404 error when trying to view my renaming list (and yours too). I can see all the others just fine. --dayna 11:43, 19 August 2009 (EDT)

Sorry about that. I had an extra "r" in your user name--it's fixed now. Also I just uploaded my list :-). Thanks for pointing it out!--Dallan 14:56, 19 August 2009 (EDT)

If only I didn't have such an unwieldy username! ;) --dayna 15:05, 19 August 2009 (EDT)

BTW, I can change it if you want, but it's not a problem the way it is either. I'm glad you don't use the umlaut though, as in Gewürztraminer. :-) --Dallan 02:41, 20 August 2009 (EDT)

If you can change it to "dayna" without any problems or hassle, that would be great! --dayna 09:10, 20 August 2009 (EDT)

The first letter has to be capitalized (so "Dayna"). Is that ok?--Dallan 12:27, 20 August 2009 (EDT)

Absolutely! Thanks! --dayna 18:41, 20 August 2009 (EDT)

Done. Your user name is now User:Dayna and your old user page redirects there.--Dallan 09:08, 22 August 2009 (EDT)

"Other" lists [19 August 2009]

I get a 404 error trying to get to any of the "Other" lists. --Ajcrow 15:00, 19 August 2009 (EDT)

Same here. --Mike (mksmith) 16:36, 19 August 2009 (EDT)

Argh. I should have tested the links. They're fixed now. Thanks.--Dallan 16:38, 19 August 2009 (EDT)

Thanks, Dallan! --Ajcrow 18:41, 19 August 2009 (EDT)

I'm on board! [19 August 2009]

Just wanted to let you know that I'm on board and will get to my list later today/tonight. Right now I'm trying to find a job. jillaine 15:01, 19 August 2009 (EDT)

Thanks! Good luck with finding a job.--Dallan 16:40, 19 August 2009 (EDT)

Census pages [20 August 2009]

The census Source pages on my list (of which there are a lot) all appear in the following form:

Old form: Source:United States, Tennessee, Campbell. 1850 U.S. Census Population Schedule

New form: Campbell, Tennessee, United States. United States, Tennessee, Campbell. 1850 U.S. Census Population Schedule

Maybe I'm missing something obvious -- but why would the place hierarchy be duplicated in the new title line? Once "up" and once "down"? --Mike (mksmith) 16:44, 19 August 2009 (EDT)

It appears that on every source that I have titled using the (old) new geographic place rule, the proposed new title includes small-to-large and large-to-small place designation. Are we supposed to change them now? I though you told Amelia the system would do it. --Judy (jlanoux) 17:55, 19 August 2009 (EDT)

I am supposed to be on vacation, but this is fine, gives me something to do while hubby is sleeping. <g> Just tell me if my changes are correct to this source page: Source:United States, Texas, Aransas. 1930 U.S. Census Population Schedule --Beth 20:26, 19 August 2009 (EDT)

Beth, I think the problem is that the bot is taking information from more than one text field on the page and combining the two. If that's the case, there's way too many standard-form census pages to change by hand. Dallan will have to adjust the algorithm to do them all at once -- and we don't know which field he's going to pick as the source of the page title info, so we don't know which one to delete the current info from. --Mike (mksmith) 21:57, 19 August 2009 (EDT)

We do know what field he's using - place. You correct double places by making sure the place covered is correct (small to large order) and that the title does not include the place. You leave the ones that have large to small in the current title and small to large in the new title as they are, since there isn't a way to "fix" them - that's how it's supposed to be.--Amelia 00:21, 20 August 2009 (EDT)

Beth, your change to Source:United States, Texas, Aransas. 1930 U.S. Census Population Schedule is perfect. As Amelia suggests, if the source title field contains the place covered, it needs to be removed, which you did. (Making sure the place covered is correct is also very important.)

I'll write a program tomorrow to see how many sources have a source title field that starts with "United States..." and see if I can remove it automatically or ask Taylor to remove it since it seems to be a common problem.--Dallan 02:41, 20 August 2009 (EDT)

Yes, but now you are limiting the places covered to only the county; I can no longer include various cities. Unfortunately. --Beth 17:12, 20 August 2009 (EDT)

You should be able to put cities in the places-covered field. If that's not working somewhere, would you give me an example?

I did not mean that this was not working; but if you are using the places covered field in the automatic changes of sources, I assume that having town in the places covered would mess that up. Is that correct or not? I thought it was helpful to have the places linked under what links here.--Beth 18:16, 20 August 2009 (EDT)

If the source covers just one town, then you should list that town in the place covered field. If it covers multiple towns, list them all, but list the primary town first. If the source covers an entire county, list the county first. You can also list the towns covered under the county in the place covered field, but it isn't strictly necessary. Sources of type "Government/Church records" will use the first place-covered as part of the source page title. Other types of sources ignore the place covered field when constructing the source page title.--Dallan 00:13, 21 August 2009 (EDT)

If the source covers the entire county, you don't need to enter each of the cities in the county in the place-covered field. Entering the county is sufficient.--Dallan 17:48, 20 August 2009 (EDT)

Problem where middle initial is 'A' [20 August 2009]

For Source:White, Stephen A. Dictionnaire Généalogique des Familles Acadiennes, Premiére Partie, 1636 à 1714 the author field on the page looks correct, but the proposed change has his middle initial changed to 'a'.
Should I make a note to change it manually after conversion?--Judy (jlanoux) 18:01, 19 August 2009 (EDT)

I should have clarified that this seems to be an 'A' problem. Other initials translated correctly. --Judy (jlanoux) 18:32, 19 August 2009 (EDT)

I know why this is happening. I'm sending the author field through the same capitalization routine that the source title goes through, which lowercases short words like "a". I'll fix this tomorrow. Thanks for pointing it out.--Dallan 02:41, 20 August 2009 (EDT)

Source page titles for websites [21 August 2009]

Looking at the table of proposed styles, something occurs to me, . . . which probably I should have thought of earlier, actually. A lot of websites (and I visit and write reviews on a couple hundred of them every year, for publication) have titles that are basically indistinguishable from books or other traditional publications. "Descendants of Valentine McGill" comes to mind; that's the name of a website, and there's also a book of exactly the same title. But it's hard to tell which type you might be looking at when scanning a list of search results. I'd like to suggest that websites are different enough from other types of sources to require slightly "special" handling, by adding an identifier of some kind in parentheses after the page title: "(Online)" or "(Web)" or "(Website)," or something similar. This assumes the robot can correctly identify websites when it goes through the renaming process, however. --Mike (mksmith) 18:41, 19 August 2009 (EDT)

What if I added source type to the search results list instead? (Website is a new source type.) Or I could add "(website)" to the end of the source page title for sources with a type of "Website") . What do others think?--Dallan 02:41, 20 August 2009 (EDT)

I would think that adding "source type" to the search results list would be appropriate, rather than adding "(Website)" or "(Book)" or "(Article)" to the source title. If you have a source titled, "Source:Smith, John. History of the World" it could be a book, article, website, manuscript and would be clearly indicated as such in the "source type" block on the source page. --BobC 12:30, 20 August 2009 (EDT)

I agree. Having "source type" appear in a column (or just followed-on in the text for the item) in the search-results list would work fine, I think, in helping the user to distinguish the different types that may have similar titles. And it would require less "fixing." too. --Mike (mksmith) 16:33, 20 August 2009 (EDT)

I do not care in the slightest whether the system (or even a human editor) thinks something is a "book" or a "website" given how fine the distinction is these days. Understanding that some people do, adding it to the search results is fine, but please do not clutter up the titles with any more information!--Amelia 11:07, 21 August 2009 (EDT)

Foreign stuff [21 August 2009]

Am I correct that the limit is three levels in the hierarchy of a place-name in a page title? Or maybe it's four levels, if you include a town name. However, on my list is Source:Registres_de_Bezannes, a church register, which had only the town name as a place issued/covered. When I added "Marne, France" in the places-covered box, the system changed it to "Bezannes, Châlons-sur-Marne (arrondissement), Marne, Champagne, France" -- which is five levels, plus a qualifier. Do we really want something like that in the page title? But you can't really just truncate it, either, and have it make sense. --Mike (mksmith) 19:11, 19 August 2009 (EDT)

And I've found several others on the list with the same issue. --Mike (mksmith) 20:47, 19 August 2009 (EDT)

Old form: Source:Conrad, Glenn R. Saint-Jean-Baptiste des Allemands
New form: Conrad, Glenn R. Saint-Jean-Baptiste Des Allemands

You need to add prepositions in foreign languages, at least the common ones, to your "do not capitalize" list in the capitalization algorithm. In French that would include le, la, les, de, des". --Mike (mksmith) 19:30, 19 August 2009 (EDT)

And I've found several where "through" (a preposition) is erroneously capitalized in English-language titles. --Mike (mksmith) 20:58, 19 August 2009 (EDT)

I'll add the additional prepositions. By the way, if someone could point me to a good list of foreign-language prepositions or could write down prepositions for German and Dutch if you know them off the top of your head I'd appreciate it.

I use Cindex for indexing books, which happens to have a convenient list of English prepositions (which one doesn't cap) in its "sort" function, which is this: a against and as at before between by during for from in into of on the to under versus with within. In French and German it's a bit more problematic, since they don't necessarily follow the English language capitalization rules. The French list (much shorter) is above. Judy's the German-speaker in the family and she can give you that. Dutch would be very similar. "Hey, guys! Anyone here speak Spanish?" --Mike (mksmith) 09:47, 20 August 2009 (EDT)

I found a quick list of French simple prepositions: à après avant avec contre dans de depuis dévant des dès durant en jusque les par passé pendant pour près sans suivant sur vers. French articles are le la on les un une.

Those should cover 99.99% of the instances. --Mike (mksmith) 10:05, 20 August 2009 (EDT)

I'd read somewhere that prepositions with five or more letters should not be capitalized, but looking this morning it looks like other sources say to capitalize them. So the list of words not to capitalize will be (words I had previously from other lists are in italics):

a, against, an, and, as, at, before, between, but, by, during, for, from, in, into, nor, of, on, or, over, per, the, through, to, under, up, upon, versus, vs, with, within

plus the french words you listed,

plus the following spanish words (may not be a complete list):

contra, y, ante, antes, entre, pero, por, durante, para, en, o, sobre, el, mediante, bajo, con

Your list above didn't include the word "through" but I added it as well.--Dallan 12:23, 20 August 2009 (EDT)

This is causing a problem with the Source History of La Porte County, Indiana. The proposed renaming makes the La part of La Porte lower case. This just happens to be the county (and city) in which I live, so I noticed it right away. --dayna 18:37, 20 August 2009 (EDT)

Here's a WP link for Dutch http://en.wiktionary.org/wiki/Category:Dutch_prepositions

Working on German...It's not just a matter of translation - the usage is different. Germans capitalize a lot more than English. I would start with the 5 letter rule you mentioned. I can still recite "an auf hinter in neben uber unter vor zwischen" from my college days. The other lists are harder...

Contractions (article + prep): am ans aufs beim furs im ins vom zum
dative preps: aus ausser (long s) bei binnen entgegen gegenuber gamass (long s) mit nach nebst samt seit von zu zuwider
accusative preps: durch entlang fur gegen ohne um wider
All assume we are omitting umlauts and using 'ss' for long s. Articles are: der die das ein - but they have many variations.

--Judy (jlanoux) 13:07, 20 August 2009 (EDT)

I read somewhere too that Germans captitalize a lot more. So do we want to lowercase all of the above, or a subset of them? I'll definitly add an auf hinter in neben uber unter vor zwischen der die das ein am ans aufs beim furs im ins vom zum. My question is about the dative and accusative preps. Also, can you tell me which u's could have umlauts?--Dallan 14:24, 20 August 2009 (EDT)

I was going to study some German book lists for implied rules, but then we'd never get on with this project. Can't we have a cleanup run later and just go with a few for now?

Umlauted u in uber gegenuber fur furs

Umlauted a in gemass

--Judy (jlanoux) 15:45, 20 August 2009 (EDT)

A later cleanup sounds like a good idea. I spent too much time myself on the spanish prepositions this morning. I'll add gegenüber, für, fürs, and gemäss to the lowercase-words list, so they'll take effect starting tomorrow. I'll add only the umlautted forms of fur and furs though, since I don't want those words accidentally being lowercased in English-language titles.--Dallan 17:42, 20 August 2009 (EDT)

Regarding place hierarchies, most countries have 3 or 4 levels. Some places in Germany have 5. France is the exception with six levels in their place hierarchy: commune, canton, arrondissement, department, region, France. The problem is I don't know enough about French genealogy to know which level(s) I can omit. I'd say that ideally we'd pare the place hierarchy in France down to 4 levels, update the place wiki to remove the additional levels, and title the sources using a 4-level place hierarchy like the US. The FHLC has some records stored at all six levels, but most appear to be at the department and commune levels. Would we be ok locating communes, cantons, and arrondissements all directly under departments? I'm open to suggestions.--Dallan 02:41, 20 August 2009 (EDT)

I looked at French govenmental & tourism sources, and also at Wikipedia articles on French communes when I was first creating those sources. I'm afraid I don't see any obvious guides as to which levels in the hierarchy are routinely omitted by the French themselves. (But then, you know the French.) I would suggest, looking at the way places are described in English-language books (not necessarily genealogy), that you would have to include these: commune, arrondissement, department, France. That locates the place precisely enough to find it on a map (including the problem of duplicated commune names, of which there are a great many across France). I think I would call up the local college and ask someone in the French department, preferably someone younger, who travels in France in the summers. --Mike (mksmith) 10:12, 20 August 2009 (EDT)

That's a good idea. In fact, I may be able to talk with someone who also does French genealogy. In the meantime, let's be consistent with French sources and put all levels in the title. Once I get French places straightened out it won't be difficult to automatically rename the French sources.--Dallan 12:26, 20 August 2009 (EDT)

A title with "A.D. 1800" got changed to "a.D. 1800", apparently mistaking the A. for an article. Just FYI, I'm sure somebody'll fix it later anyway. --Jrich 20:57, 20 August 2009 (EDT)

I'm going through the titles right now reviewing the foreign words that are in the to-lowercase list to see what kinds of titles they tend to be in, and if they should be lowercased or uppercased the majority of the time. We're not going to get the page title casing right every time -- I probably won't be able to automatically fix the A.D. situation, but hopefully we'll make the right choice in the majority of cases. Ultimately the source title field is what will appear in the text, not the source page title, so hopefully a few mistakes won't be that big of a deal.--Dallan 22:51, 20 August 2009 (EDT)

After looking at titles with foreign articles in them, I think we're going to have to deal with having badly-capitalized source page titles when the title includes place and person names that have foreign articles. So for example, "La Porte" is going to appear in the source page title as "la Porte". The source page title won't be used for generating source citations though; we'll use the title field within the source page for that. We're standardizing the capitalization on source page titles mainly to reduce the chance of creating duplicate sources with the same title but different capitalization. The source page titles can be renamed later if you want.

The foreign articles de, des, el, la, and le are the only foreign words that we're not capitalizing that appear frequently in titles. They each currently appear far more often not capitalized (because they're used as articles) than capitalized (because they're part of person/place names).

I'm starting to think that maybe changing the capitalization of the current title is a bad idea. Instead, the system could just copy the source title field into the source page title as-is. Most of the source title fields capitalize just the first word and proper nouns. How would people feel about this -- not changing the capitalization of the current title? We'd likely end up with some inconsistency in titling because some people would create sources with most words capitalized, while others would create sources with most words not, but we wouldn't have weird capitalizations like "la Porte" and "des Moines".--Dallan 11:37, 21 August 2009 (EDT)

I've been wondering why we're bothering with this. My vote is to leave it alone. We're likely to cause more problems than we fix.--Judy (jlanoux) 12:42, 21 August 2009 (EDT)

So you're proposing a major style change that affects the vast majority of the human edited sources, thus making there be no rhyme or reason in the source database, in a bottom comment of a thread on another topic? In the interest of sparing the dozen Des Moines sources from embarrassment of mis-capitalization? I would rather have an actual rule that people can follow, and fix the goofy ones accordingly when they show up.--Amelia 14:08, 21 August 2009 (EDT)

You know, I've spent so much time lately looking at the FHLC sources that I forgot about the several thousand human-created sources that have been capitalized by the system when I wrote my message. I agree that we don't want them to look odd. I just had an idea: For the renaming if a word that we would normally lowercase is currently uppercased in the source title field, then we'll keep it uppercased. Going forward, we'll lowercase words like "des" in "Des Moines" when people add new sources unless they uppercase the word when they enter the title on the add source screen, in which case we'll keep it uppercased in the source page title. Thanks for the reminder.--Dallan 15:26, 21 August 2009 (EDT)

Newspapers [20 August 2009]

Source:The Bloomfield News, published in Bloomfield, Indiana, will have its article omitted -- "Bloomfield News." Good enough. But the new form will be "Bloomfield News (Bloomfield, Greene, Indiana, United States-Bloomfield, Greene County, Indiana)" -- taking info, it looks like, from both the places-issued box and the places-covered box. --Mike (mksmith) 19:20, 19 August 2009 (EDT)

Yes -- I've found several others on the list in which the parenthetical info follows the same pattern of duplication. --Mike (mksmith) 20:54, 19 August 2009 (EDT)

The place issued field is causing the problem:

• Until very recently the place issued field got standardized and all levels in the place hierarchy were added. I've stopped doing this, but a lot of place issued fields have unnecessary levels. In my opinion, "Bloomfield, IN" is probably sufficient for a place issued, not "Bloomfield, Greene, Indiana, United States". If you wanted, you could edit the place issued field and replace the text with just "Bloomfield, IN" (it doesn't get standardized anymore so it will take exactly what you enter). Then the source title would be "Bloomfield News (Bloomfield, IN)".

• Another problem is that if the place issued has a bar(|), as in "Bloomfield, Greene, Indiana, United States|Bloomfield, Greene County, Indiana" I'm taking the entire text of the field, converting the bar to a dash, and adding it to the title. I should instead take just what appears after the bar, so in the above case if you didn't edit the place issued field, the source title should be: "Bloomfield News (Bloomfield, Greene County, Indiana)". I'll make this change tomorrow. Thanks for pointing it out.

--Dallan 02:41, 20 August 2009 (EDT)

For bibliographers and librarians and such, the rule of thumb for newspapers is, if the place-name is part of the name of the paper, and it's a large city -- New York Times, Toledo Blade -- you don't have to do anything. If the place-name is included, but it's a smaller place that might be duplicated elsewhere -- Bloomfield News, Denton Ledger -- you add the state in brackets, for clarity: Bloomfield [Indiana] News, Denton [Texas] Ledger. If there's no place-name in the name of the paper at all -- The Advocate -- you add the town and the state at the end: The Advocate [Baton Rouge, Louisiana], The Advocate [Victoria, Texas]. All of this is because (of course) there are a relatively few newspaper names in heavy and widespread use, and you have to distinguish among them.

For WeRelate, though, I think it could be boiled down to a simple yes-or-no rule, something like: "If the place-name is included in the newspaper name, don't do anything. If it isn't, always add the city and state." That would have to be applied in each case by whomever creates the source. Even simpler would be: "Always include the city and state following the name of the paper," period. That could be easily robotized, I think. Having laid out the theory, though, now that I know where the extra place names are coming from, I'll go back to the sources I created, at least, and say if I can't tidy them up. --Mike (mksmith) 17:14, 20 August 2009 (EDT)

To keep things simple, I'm following the simplest "always include the city and state following the name of of the paper" rule. As of now if there's a bar(|) in the place issued I'm taking whatever follows the bar so hopefully the newspaper sources that you were looking at yesterday are improved (maybe not perfect). Since we're no longer trying to standardize place issued, we shouldn't have this problem going forward.--Dallan 17:53, 20 August 2009 (EDT)

Acronyms [20 August 2009]

The system seems to be taking acronyms and changing from all caps to initial caps only. Example: The AHSGR Passenger Ship List Index --Judy (jlanoux) 21:13, 19 August 2009 (EDT)

Ok, I'll leave words in all-caps alone.--Dallan 02:41, 20 August 2009 (EDT)

FHL Numbers in title [20 August 2009]

Hello,

Are still deleting the FHL numbers in the titles, or are we moving them to somewhere else in the web page? Thanks Debbie Freeman --DFree 22:00, 19 August 2009 (EDT)

We're deleting the FHL numbers in the titles. The web pages will still have links to the proper page in the FHLC, so I don't think we're losing anything by deleting the FHL number from the title.

Having said this, there are currently 50,000 sources that will end up with the same title as another source after the renaming. Most of these sources are FHLC sources for government/church records that have exactly the same title and place-covered fields. Often the only difference between these sources is that the author (person/organization compiling the records?) will differ. Rather than try to resolve this situation (many of the sources are in languages I can't begin to read), I'm planning to keep the FHL number in the title for those sources where taking it out would mean that it's title would duplicate that of another source.--Dallan 02:41, 20 August 2009 (EDT)

--- Hello Dallan, As for the same name sources could we add more types such as online book, original textual record, on line database, etc? Could that make a source only one page/source, but choosing a type depending on what the Users have access to? Would that solve the user's same name search problem, etc? Debbie Freeman --DFree 17:50, 20 August 2009 (EDT)

I've been thinking about this too -- the possibility of having subpages for different locations of a source. I'm going to post something about this in the topic below. But that's not going to solve the duplication issue. The main problem there seems to be when we have two distinct sources that just happen to have the same page title due to poorly-specified titles and/or authors. I'll hopefully have some examples to show tomorrow, but I think I'd like to wait to deal with these sources until after everything else gets reviewed.--Dallan 00:07, 21 August 2009 (EDT)

Government/Church records rule [22 August 2009]

It seems like the statement of the policy for naming websites isn't consistent with the one for books. If an individual or society transcribes records into a book, we cite the book (type = book). So if someone posts records on a website, shouldn't we cite the website (type = website) to retain the distinction between the official record and the private copy? I would agree, however, that official government websites can be considered Government records type. --Judy (jlanoux) 23:54, 19 August 2009 (EDT)

This is a really good question. On the one hand, I like the idea of having a source page titled Amador, California, United States. Death Index, Pre-1905 for example, because it allows people to add additional repositories for this information later. On the other hand, it is different from how we're treating books. Amelia would say that we should treat be treating books containing government/church records as having a source type of "Government/Church records", not book. Perhaps she's right. It would sure be nice to have a simple, easily-understood rule for when to treat a compilation of government/church records as "Government/Church records" and when to treat it as an authored "Book".

This is a complicated topic. For instance the Amador project is part of a statewide project, but the index is created and transcribed by volunteers from county records. See this link [[1]].

I guess for me it boils down to what we eventually want to have in the generated source citation: do we want to cite the website, or do we want to cite the original data? For an official government website, it's easy: cite the original data, so set the source type to "Government/Church records". For rootsweb transcriptions, I lean toward citing the original data (so set the source type to "Government/Church records"), but maybe it would be better to cite the website instead.--Dallan 02:41, 20 August 2009 (EDT)

I would argue that in instances where you have transcriptions on a website, you cite the website. Two reasons: (1) why should we treat transcriptions on a website different than transcriptions published in a book (which has the citation to the book, not the original source) and (2) although there is a difference between a Source Page title and a source citation/reference/footnote, a lot of users are not going to make that distinction. --Ajcrow 07:08, 20 August 2009 (EDT)

I agree absolutely on this (and so does Elizabeth Mills). You cite whatever you actually looked at. If you saw a transcription of a record on a website, which copied it from a book, which took it from the original records in the courthouse, you cite the website. We do a lot of courthouse-prowling and I always make a distinction between the original record that I eyeballed in person and took a digital photo of, and the published transcription (or, worse, an abstract) that led me to the original. --Mike (mksmith) 17:19, 20 August 2009 (EDT)

It is especially because you plan to have real citations that I pointed this out. Here I do disagree with Amelia. Anybody's copy is not the same source as the original records and cannot be cited as such. I think we have the subject field to show what the contents are (would like a few more choices there) and don't need to use the type field for this.

Where necessary to guarantee uniqueness I favor adding the word "site" to the title indicate a website, as in "XYZ County Marriages Site". --Judy (jlanoux) 09:41, 20 August 2009 (EDT)

Maybe I'm just brain-dead today, as I have lots of "real world" projects with an immediate deadline right now, but I remain very confused by this discussion. The example on my list is Source:United States, Massachusetts, Norfolk, Braintree. Vital Records and Church Records of Braintree, Massachusetts. This is not the official Braintree MA vital records; it is a lifelong compilation (index cards) by a genealogist from public records, church records and "private sources" (read: Bible records probably) that was donated the NEHGS. It is available in their manuscript archives, and has been turned into an online database by NEHGS, available on New England Ancestors.org. Is this a "website" source? A "government/church" records source? I chose government/church records, as I think that is how it should be titled (with the geography in front of the title).

My premise is that the SOURCE is (primarily) government/church records. The source data is extracted on a website, and can be found in the NEHGS library. But the website per se isn't (exactly) the source. I would think most (tho not all) websites are digitizations of real world sources, be it a book or vital records or.... I would like to know what the original source is for information, as well as the location of a version of that information. I think of a web page transcription as almost like another "edition" of a book or of vital records.

Caveats are needed, as a transcription or a website or a database can't be guaranteed to be complete and accurate representations of the original sources.

On the other hand, in this case, pretty much the only realistic access to this particular compilation of vital records is probably through the NEHGS website (or archive), as it is an individual's body of work. So maybe it is better described as a website source?

Maybe there also needs to be a way to distinguish between a source *Database* (website based, common on Ancestry, and also NewEnglandAncestors), based on extracted data and names from a source or a variety of sources, and a book or vital records or other source type (whether available on the web or in real life or both). Otherwise, won't we end up with duplicate sources, one for the "book" version and possible multiple other sources that are website variations of the book?

So I've come full circle to thinking that perhaps my Braintree Vital and Church records should be a source type "website," meaning a unique online-only database extraction of the original records. But maybe also the source type "website" should be renamed to clarify? - brenda--kennebec1 11:21, 20 August 2009 (EDT)

I agonized over one of these last night. I tend to want it to be cited like a book and credited to the author. Mike objects to calling it a book if it wasn't actually published and bound but I'm a little more pragmatic and can deal with an "unpublished book". Is this what is meant by the "manuscript" type? - someone's unpublished notes? I always thought that was for things like collections of letters and papers, but could be wrong. I think I went with "book" as that was the one that gives a proper citation. The website is a repository just as any other website that has books on it (google, ancestry, etc.). While it is tempting to take the easy way out and call it government records it really isn't correct to cite it that way. --Judy (jlanoux) 12:30, 20 August 2009 (EDT)

It's the old problem: How many steps removed are you from the creator of the record? To me, a microfilm of a census page is equivalent to the original bound ledger because it's an image, and likewise with the digitized census page images on Ancestry. But a transcription of a cemetery grave marker is one step removed -- and the copy of that transcription that someone later types into a website is two steps removed. Each "step removed" means added questions about accuracy and completeness. That's why you don't say a published book of collected records is the same as the original records. --Mike (mksmith) 17:28, 20 August 2009 (EDT)

As I've thought about this more this morning, I have a couple of observations:

• Books are long-lasting, websites are not.
• Data on websites changes over time. This includes the website title.
• It's often difficult to determine the author of the website.
• Many websites have poor (not-fully-descriptive) titles.

The above can lead to the following bad situations, which will increase in frequency over time:

• We'll end up with a lot of different websites with the same titles
• We'll end up with multiple sources for the same website as it changes its title over time
• We'll end up with sources for websites with non-descriptive titles where the URL is no longer valid. How can people tell what the person who added the source originally was trying to cite?

I think we agree that an Ancestry, HeritageQuest, and now FamilySearch (record search) online censuses need to share the same Source page. We'll have the situation where someone citing that source wants to specify exactly where they looked up the information. To do this we'll have to give them a way to add the repository they used to the source citation on the person/family page. Then when we generate the source citation, we would need to include the information from their citation as well as from the source page itself. So we'll have to have a way in source citations on person/family pages to specify the repository or url where the information was found.

So I'm proposing that in order to avoid creating a bunch of sources for websites that are likely to become a problem to maintain over time:

• if the website is a transcription/compilation of an offline government/church record set (a set of records of the same type, covering a particular locality and time period), then we give it a source type of government/church record set, and we give people a way in their citations to specify the date and place they retrieved the information from (e.g., add "as downloaded from < url > on < date >" to the generated citation)
• otherwise if the website is a unique collection of material, say an aggregate of different types of records possibly covering multiple localities and time periods, that we give it a source type of website.

Then if the website goes away or changes its title and/or url over time, at least people will know the type of records that were cited and they might even see another repository for the records listed on the source page, and people who want to be careful can add information to their citations on person/family pages to cause reasonable citations to be generated.

We might even say something similar (but slightly different) for books:

• if the book is a transcription/compilation of a government/church record set that was put together by a government agency, then we give it a source type of government/church record set
• if the book was created by a human being or non-governmental organization, then it's a book.

This rule is different from the website rule because books are long-lasting, have stable titles and relatively easy-to-find authors, and the data in a particular edition doesn't change over time, so book citations are fairly persistent.

Thoughts?--Dallan 13:54, 20 August 2009 (EDT)

For books, I think your distinction between published documents prepared by the same agency that created the orignals, vs. third-party compilations, is spot-on. I'm thinking about all the published records surveyed by the WPA. And all the post-Civil War "registers of veterans" published by state adjutants-general. And the published American State Papers, for that matter. Judy's right: I'm a book person. Calling a non-hard-copy collection (e.g., a website) a "book," just to get a citation that includes the fields you want, seems almost . . . immoral. (Sorry.) But the distinction here is a good one. --Mike (mksmith) 17:40, 20 August 2009 (EDT)

The responsibility for a correct citation ultimately rests on the user. We can't provide a canned citation that is complete. So the user should be stating which version he has used (according to ESM). The problem I have had with using 3rd party secondary sources as Gov/Ch records is that the source title would be indistinguishable from the original. We would have duplicate titles. If we make the web site one of multiple repositories and allow the user to select, we would be encouraging good practices. But if the type is Gov/Ch then I would not want a 3rd party copy to be called "the source". This proposal works for me with websites. But can I have Courthouse and Archives as Repository types, please? We need to offer the originals as one of the choices.

Our catalog seems to not have entries for films of the original courthouse or church records - that I have found - even though I know many of them were filmed. So it appears that the website entries are defining the availability of records. The problem with implementing your proposal is "What are we going to use as the title?" For the original records I have always favored the form that has been promoted since the 70s - Dallas County, Iowa Marriages. If we use our place convention and the recordset name - Dallas, Iowa, United States. Marriages - I would think we're pretty darn close, hooray! But if we use - Dallas, Iowa, United States. Marriages 1901 - because that's the ones someone stuck on a rootsweb page we're back to letting the 3rd party masquerade as an official record. I would suggest using the text box to describe what each repository includes and naming the page more generically. It would be very useful to have all the transcriptions gathered up into one page. I would like that.

Is this project getting bigger as we go - seems like it? Are you going to create that marker type for inappropriate sources? --Judy (jlanoux) 15:38, 20 August 2009 (EDT)

I agree that the responsibility for a correct citation lies with the user. I just have a hard time labeling a website with record abstracts as Government/Church the same as we would the original (or microfilmed or scanned) records. One thing that could help would be a note on the Source page that the website is abstracts only. Thoughts? --Amy (Ajcrow) 16:53, 20 August 2009 (EDT)

One item that has not being considered is how to simplify this so new users may understand it easily. For heavens sake, it is no fun to have to look up 10 different things every time you want to enter a person, events, and citations. I can't even remember all of it. --Beth 18:26, 20 August 2009 (EDT)

I'm okay with Dallan's proposal. As a corollary to the 'responsibility for a correct citation lies with the user', I'd like to point out that these citations are not user-specific. They are for the community. As such, we cannot and should not be trying to generate citations that reflect "where I saw it" in a particular library/repository/website -- there is no relevant "I". A particular user can document his observations in the source notes or on the source page, verifying that the cited information is in some version or another, but since the criteria for things in different repositories sharing a source page is that they be functionally the same, it should be really rare where data is missing from one version that appears in another. With this in mind, yes, website transcriptions of vital records show up only as repositories to the record set source. Nothing on NEHGS, Ancestry, or Heritage Quest should have its own page. The "website" isn't a source, so don't worry about it being "the same" as a site for the "original". Only the "original" gets a source page. Even if we can't figure out what the original is right now, precisely, creating a general source page will put that knowledge out there and allow collaborators in a way that isolating it as a "website" source won't. --Amelia 11:04, 21 August 2009 (EDT)

The "where I saw it" under discussion is the original document (which includes an actual image of the original document, as presented on Ancestry or wherever) vs. some person's published transcription of the original document. This is not a user-specific issue. The genealogical community decided long ago that the original and a transcription of the original -- whether in a book or on a website -- are not equivalent. You need a source for the original documents in a county clerk's office, and you need a separate source for some compiler's rendering of what those originals said. To assume that the two are the same is very bad research practice. --Mike (mksmith) 13:12, 21 August 2009 (EDT)

Do we have a solution for this yet? I have quite a few of these on the list I'm reviewing. -- Amy (Ajcrow) 08:51, 21 August 2009 (EDT)

Please update the project page once a decision is made to change something so we know which way to jump. I've been thinking about this issue as I work my list. We have two fields: Type and Subject. Type seems to be indicative of the media, subject to reflect the contents. So it is feasible to leave the type as website and use the Subject to decide if we should use geographic format? Then we would have a label like: "County, State, Contry". Jolly Joes Marriage List. I could be happy with that. --Judy (jlanoux) 12:19, 21 August 2009 (EDT)

I *really* like that solution, Judy. -- Amy (Ajcrow) 14:41, 21 August 2009 (EDT)

Here's an idea for a solution that I think may address the issue. Going through ESM's book I notice that citations generally contain two parts: what you found and where you found it. When the system tries to generate a citation from a link to a source page, it can get the what from the source page fields, but not the where, because it won't know which repository you looked at. Here are two alternatives for capturing the where information:

1. When you link to a source page, you're prompted to select which of the repositories listed on the source page you looked at. If you select a repository (you don't have to), the information from the selected repository is copied to your citation reference in the person/family page. The full text citation is generated from the fields on the Source page and the repository fields that have been copied into your citation on the person/family page.
2. We create subpages for each repository or location at which the source can be found. The subpage title would be the title of the source page, followed by a "/", followed by a short word or phrase to identify the location, like "Dallas, Iowa, United States. Marriages/Rootsweb". It just needs to identify which of the (handful of) locations this source can be found at. Most sources would have at least one subpage, created automatically from the repository information. The subpage contains the information needed to fill in the where portion of the full text citation, like the item type/format (digital images, transcription, originals), the title of the website, URL, etc. The subpage could contain a link to a Repository page when appropriate. Instead of adding new repositories/locations for the source by editing the source page, I would put a button on the source page that would take you to an add-page screen for adding the subpage. The drop-down list would include all of the subpages and would include source pages only for those sources that don't list any repositories.

The advantage of option 2 is that if the website title or URL changes, someone can update the subpage and all of the full text citations stay in sync. Also, it's possible to see who is citing a particular location because you can do a "what links here" on the subpage. The disadvantage is that the drop-down list is larger and we have more pages overall to deal with. I probably wouldn't want to index the subpages, because I'm not sure it makes sense to have the subpages showing up in search results (or we should at least hide them during most searches).

If we went with option 2, I would remove the repository information from the source pages and use it to generate the subpages automatically. And I would modify the system to rename source subpages automatically when the source page is renamed. But I don't plan to implement either option until after the big renaming project is complete.

Either of these approaches will allow us to generate citations that cite the website, not the original records. Given this, and because of the difficulty that I think we'll have with maintaining Source pages with website titles long-term, I think we should prefer the "government/church records" source type for websites, unless the website contains a hodge-podge of material that can't reasonably be thought of as a cohesive government/church record set; i.e., contains a variety of different types of records. Only in these cases should we call it a website.

Sound ok?--Dallan 12:42, 21 August 2009 (EDT)

Judy and I have talking/arguing about this and have been tending toward something like your No. 2. I think this could be made to work. And I think you're making an unstated but entirely reasonable assumption about something there which I'll make explicit: Not every Source will need repository subpages. If it's an actual book, it's immaterial which library you found it at, or even whether you found it at Google Books; the content in this case is identical. (I suppose you could have subpages for reprint editions, but that seems like overkill.)

Obviously, there are going to be an enormous number of subpages pointing to microfilm available from the FHL. Is it possible to auto-magically scoop up the film numbers and all that from the text box and move them to the FHL repository-subpage for the Source? Anyway, the microfilm is equivalent to the original records, because they're images. But a transcription on somebody's website is not equivalent to the original, so we come around to that again. . . . --Mike (mksmith) 13:41, 21 August 2009 (EDT)

To be honest, I don't understand what problem you're trying to solve, Dallan. I know that ESM says you're supposed to say where you found it, but that's for one user properly documenting that one person's work. This is a community with lots of collaborators. As Mike notes, whether I saw it in hard copy at NEHGS and you saw it at the Allen Co public library and someone else saw it digitized on Ancestry.com is irrelevant to the next user, who just needs to know that the fact is in this record set, which can be found at these repositories, where he can go look at it himself. How would Option 1 generate a useful citation in that case? What information would be captured on the Repository subpages that can't be captured on the source page? You're just making users do extra clicks for no benefit.

I think we get entirely too wrapped up in what is Proper Research Practice for an individual researcher. Our goal is to point people to a source that proves a fact, and we don't need to send people to a particular library or website to do that because the source in the vast majority of cases will be exactly the same in every repository. And my comments are based on that assumption -- if the sources are materially different, they get different pages. So as for the transcriptions point, in most cases, a county records collection is going to be a separate source than a transcription of part of it because the subset is different. But sometimes transcriptions, like with censuses, get the same source page as the original. No, this isn't perfect, but the wiki benefits we get from being able to share which parts of a transcription are bad or great outweigh the downside of material inconsistencies, which can be dealt with individually in the source comment. And while I agree that best practices involve us specifying whether we are copying a transcription or doing the transcription ourselves, that can and should be done with the existing technology by noting it in the source comment, which should include a transcription anyway. (And, frankly, this is so far from the current practices of our users that making things significantly more complicated in order to make a wiz-bang system to automatically recognize these fine distinctions seems a great waste of effort.)--Amelia 13:59, 21 August 2009 (EDT)

The problem that Dallan is trying to address is not about books of transcriptions. If it's published like a book, it's a book. And I think we can all agree that digitized images of a record as well as microfilm are the same as the original. (That's where Repository can come in handy.) What is materially different about this discussion is that an abstract or an index on a website is not the same as the original record. -- Amy (Ajcrow) 14:41, 21 August 2009 (EDT)

Gosh this is a huge topic. (But I don't have the energy to try to break it up into sub sections just now.)

So I would love to use Government/Church records as the Type for books published by towns or others that include, in essence, a transcription of the original records, but now when I select Government/Church records as a type, I no longer see author and publisher. Yet, there is clearly an author (or at least a compiler) as well as a publisher (could be a town or could be a private company). So I want back all the fields associated with Book. Otherwise, I won't get an accurate citation, I'm not really being accurate in describing the source. In the meantime, I'll use Book.

jillaine 14:04, 21 August 2009 (EDT)

Agreed this topic is out of control. But if you use book, you're drastically changing the page title, so please don't use that as your rule. If it has a human author or compiler, then it is a book, yes (according to our rules), But if it was a town or public entity, it's not, and needs to have the gov/church record type in order to get the place name source.--Amelia 14:11, 21 August 2009 (EDT)

Yes, I over-simplified before. It's a book unless a government publication. My apologies. -- Amy (Ajcrow) 15:11, 21 August 2009 (EDT)

I've added publisher and publication date as fields to the "Government/Church records" type (author was already there). I'd like to put this topic to bed. Let's go with a source type of "Government/Church records" for record sets on websites and government-authored books. We can figure out how, or if, we want to distinguish between different transcriptions later, once we actually start generating source citations. I think using subpages for cases where people feel a distinction is important will work, and allows our source page titles to remain relatively consistent and stable.--Dallan 14:59, 21 August 2009 (EDT)

Sounds like a plan. The subpages could be a real benefit. -- Amy (Ajcrow) 15:11, 21 August 2009 (EDT)

Amelia, I think I agree with the feeling with your statement. Is it really beneficial to the general user community to identify all the finer points of provenance, and is programming such provenance data a useful way to spend limited programming resources? The more complicated the approach to provenance, the more user training to enable the features to be used, and correctly so. Additionally, if I, as a WeRelate user, see that someone provided data they took directly from the original, my first thought is, what is their level of expertise in reading original records, or even in knowing enough about the provenance of the records to realize they are originals, or a copy anyway? In some ways, if that same person transcribed a record from Source:Mayflower Descendant, for example, I would have a better understanding of how reliable the data is.

But I like Dallan's idea that if something is essentially a faithful copy (e.g., FHL microfilm), or a best-faith transcription (e.g., website), to treat it as a separate repository. It indicates the source is presumably equivalent, without losing the distinction that it is not identical. But the source citation would still be the place, for example, to note when ancestry.com census "record" has an invalid transcription of the ancestry.com census "image" (or familysearch.com, or heritagequest.com, or any of the other such sources).

This idea doesn't change the idea that published versions of government/church records should still be done as books (i.e., not as a repository of the government/church records). Usually they differ in some way from the original, whether in formatting, ordering, the editor's added knowledge and assumptions, or in supplemental sources used to create the published work such as the C.R., G.R., and P.R. items from published vital records. --Jrich 15:15, 21 August 2009 (EDT)

Wow, a lot can happen while I'm taking a lunch break. To address first Amelia's concerns: We are not trying to make every cite personal. In each case a person should post the best information he has available. If someone else has better data, he would replace the existing source with the better one. This is an iterative process in research. Eventually the page has high quality data and it is not ever referencing who saw what where. But the point we are making is that if you haven't seen the original, yet you cite the original, no one knows what the original really says! And no one knows that it hasn't been consulted. This is a problem. When you post, you should only post what you know from personal experience - it isn't necessary to make this a personal reference. But to claim you have seen an original source when you've only looked at an abstract on a 3rd party's website is going to make WR useless for serious researchers. It's far worse than posting facts without sources.

Currently links to sources are merely links to a catalog where a user can get more information. Making it into a citation requires effort from the user. He has to either use the text box or recast the link with a pipe. Either provides an acceptable solution.

We hit some 'hot buttons' when it was proposed that the link would in the future be turned into a citation. Whoa! Now the link isn't just a place to go for more information. We've elevated it and given it more responsibility. And now we have to be careful that the catalog can handle this extra burden. It appears to me that we're just not ready to take this on. May I propose that we shelve the citation part of the project until we can more thoroughly examine the possibilities and impacts without the pressure of the renaming deadline? I think Dallan's ideas have real potential but they would require redesigning the catalog. If we don't rush into it, we can probably end up with something much simpler for the user to handle (fewer source pages, more choices per page).

For now, I would like to call a spade a spade and name a website a website. (and we've agreed Ancestry, HQ, et als are repositories, not websites.) We need to review a duplicate list and I don't think websites will turn out to be a big culprit. If there is a problem, the title can usually be adjusted slightly. For the user it's easier if he finds what he's looking for when he goes to link a source. He's got a website, he needs to find a website. Remember the Type field is referencing media and not content. --Judy (jlanoux) 16:15, 21 August 2009 (EDT)

Judy, I think you're exactly right. I don't disagree with your first paragraph either -- I think that's exactly how people should do things. But we don't need a new system to do it.--Amelia 16:24, 21 August 2009 (EDT)

Well I really have tried to muddle through all of this discussion; but I don't understand the problem. First, I thought we originally added the geographic location to the source page title to make it simpler for people to search. Second, when did one decide that the type only applied to "official" government and church records? I fail to see why it cannot apply to all government and church records, except those authored which appear in bound books. For government and church records you could just add a secondary type and select original/digital image, abstract, transcription, compiled index, etc.--Beth 22:36, 21 August 2009 (EDT)

After reviewing some of the websites and seeing how many have changed their titles since we first added them just a few years ago, and how poor some of those titles are, I want to avoid titling source pages using only the title of the website when possible. For websites containing government/church record transcriptions, I think adding the place covered to the front of the source page title will help people find the websites faster than having the source page title be the website title alone. Also, their titles will look similar to other government/church record sources in search results, which I think is a good thing.

Only a very few of the websites I come across are given over to transcriptions of records. Maybe 90% (really) are family lineages, which may include incidental transcriptions of or links to records, but they're mostly the equivalent of "family tree" books, with perhaps a DNA project or family photos thrown in. But they also tend to have not-very-useful site-titles. And I don't think you can repair all those by adding geographical prefixes to the Source page title. --Mike (mksmith) 10:39, 22 August 2009 (EDT)

The family tree websites of which you speak were supposed to be deleted in source review. They're not useful enough to merit the headaches over what to do with them.--Amelia 11:08, 22 August 2009 (EDT)

We could title the source page using the place covered followed by the website title (I think Judy mentioned this earlier). That's generally what I've been doing: adding a place covered and changing the source type to government/church records so that the place covered will be part of the eventual source page title. In doing this we'd naturally end up with different sources for websites than for official transcriptions, since the website title would naturally be different than the official transcription title. And the source page titles for both a website transcription and the official transcription of a given set of records would look similar, which I think is desirable because the records would be similar. And when the website changes its title, which is likely, it will be easier to find the source page for it because the title will at least start the same way.

Would everyone be ok with this approach? I could create a new type called "Government/church records - Website" if people think there's a need, or we could modify the instructions to say that for websites with government/church records, you should set the source type to "government/church records", fill in the "place covered", and set the source title field to the title of the website.--Dallan 00:50, 22 August 2009 (EDT)

I don't know if you've done this yet, but I'm completely confused as to what such a new "type" would be used for, and I've spent untold hours wrapped up in discussions about source naming. What is a normal user supposed to think? And if it's going to be the same as gov/church records, why do we care what the type is? How about making sure the help includes the notion that a private website could be gov/church records if that's what it transcribes, as I think it already does?--Amelia 11:08, 22 August 2009 (EDT)

As an aside, the idea of creating one source page per original record set listing the official repositories and creating subpages of that source for each of the unofficial transcriptions, whether they be human-authored books or websites, seems like a nice way to organize sources, but I'm definitely ok with not exploring that route during the renaming.--Dallan 00:56, 22 August 2009 (EDT)

I've thought more about it; how about if I add a "Government/Church records - Website" source type tomorrow for websit transcriptions of g/c records? Pages will be titled using place-title format just like "Government/Church records".

Sound ok?--Dallan 12:07, 22 August 2009 (EDT) (Solveig's driving ;-)

To answer Amelis's question (which I just read) about why the need for a new source type, I think it represents a reasonable compromise for the people who want to call websites with unofficial transcriptions "Wersites" (see the Braintree discussion below). I'm ok with either

1. adding the new "G/C records - Website" source type, or
2. asking people to use the regular "G/C records" source type.

But I'm not wild about using the "Website" type very often due to the short-lived nature of most of those titles. So I thought that a new source type would satisfy everyone.--Dallan 13:16, 22 August 2009 (EDT)

Scope of Cleanup? [21 August 2009]

As we go through the lists are we to do all needed cleanup, including deleting inappropriate records? In the 'other' list I'm finding Source pages for everything from the AMA Journal of Dermatology to the 1870 Census listing for the John Smith Family. There are still a lot of message boards and whole state (rather than county) level census pages. --Judy (jlanoux) 00:01, 20 August 2009 (EDT)

I've seen these in my review too. Here's my current thinking: if nothing links to a "bad" source, I delete it. If something links to it, I keep it. Someday we need to do something about these sources, but I don't want to worry about it today. I just realized that I could add a new source type of "not-a-source" so that we could find these sources again easily. What do you think?

Many of these should be one-time sources. It appears people thought they had to have a link for everything they entered and created one. My thought is that if it has no content I would delete it -- the only consequence is a red link and there's usually only one cite to it. For those containing content but inappropriate your idea to label them is a good one. Perhaps we would later email the creator that he should copy to a MySource or even have the system do it. --Judy (jlanoux) 09:48, 20 August 2009 (EDT)

That makes sense. There's no loss of information if you delete a source without any content -- the title still exists. I'll add a "Move to MySource" as a temporary source type later today. After the renaming I'll email everyone with links to these pages.--Dallan 12:31, 20 August 2009 (EDT)

I've added "Should be a MySource" as a new source type to help us keep track of sources that we need to ask the contributor to post as a MySource instead of a source.--Dallan 17:42, 20 August 2009 (EDT)

Ancestry.Com [19 August 2009]

In most cases I assume Ancestry.com is coming out of the title, but for their OneWorldTree and perhaps others it should stay in. Can the algorithm not have it change to Ancestry.Com?--Amelia 00:53, 20 August 2009 (EDT)

I'll make sure .com is kept in lowercase.--Dallan 02:41, 20 August 2009 (EDT)

Quote for the Day [20 August 2009]

Perhaps appropriate for this project ( I just looked at the extent of Other 1 list).

We ourselves feel that what we are doing is just a drop in the ocean.
But the ocean would be less because of that missing drop.
Mother Teresa

Jillaine's Questions [20 August 2009]

(BTW, Can I just say what a nice job you did, Dallan et al, in assigning sources to us?!)

I don't know if that's a sarcastic comment or not! :-)

It's not sarcastic at all; most of those on my list are ones I'm interested in. jillaine 21:12, 20 August 2009 (EDT)

• Is Source:Early Massachusetts marriages prior to 1800 a book or Government/church records?

Ditto this question too... see my comments regarding "Braintree church and vital records.-brenda--kennebec1 12:03, 20 August 2009 (EDT)

Let's have this discussion in the section above.

• When a book is a "Genealogy of So and So" is its SUBJECT field a "Family Tree," a "History" or a "Manuscripts/Documents" or something else?

Ditto this question... this type of book is often defaulted as history based on the FHL import, but I *think* they may be a family tree (meaning, a genealogy text, not a place history). By the way, I've taken to calling this type of writing a "genea-graphy." (not a biography, not just a family tree, but a merger of the two...) - brenda --kennebec1 12:03, 20 August 2009 (EDT)

I'd call it a family tree, since it's closer to that. I'll change that label to say "Family tree / history".

• Will a type=Manuscript Collection use the Author. Title rule?

They will if they have an author. I'll add author as a possible field to manuscript collections later today.

• Source:Burt, Henry Martyn. The First Century of the History of Springfield. Two questions: a) is this a book or vital records (Type)? b) is this a History or Directory of Records (Subject)?

I browsed the book on Google Books, and this is not just a transcription of government records. It has a lot of textual history as well. So I'd say it's definitely a book. I'd probably give it a subject of history, although you could call it a directory as well - your call.

• History of Jerseyville, Illinois, 1822 to 1901 is a book with an author. Should it be Author. Title or geographically driven? And if the latter, how edit the page such that the robot renames it geographically?

I'd leave it as a book. Only the "government/church records" source type is geographically-driven.--Dallan 12:51, 20 August 2009 (EDT)

-- jillaine 11:28, 20 August 2009 (EDT)

Wow, some of these Page Titles are going to be LONG! [20 August 2009]

See Source:Memorial to the Pioneer Women of the Western Reserve. When the name of the author is added, it's REALLY going to be long. jillaine 21:13, 20 August 2009 (EDT)

Yes. There's a cut-off at 150 characters, and thousands of source titles are longer than that and will be cut off. But what can you do?--Dallan 22:56, 20 August 2009 (EDT)

Were the lists updated? [20 August 2009]

I'm not seeing effects of the changes I made last night. I thought it said the lists would be refreshed every day so we could check if our changes worked.--Judy (jlanoux) 13:13, 20 August 2009 (EDT)

Not yet. I wanted to make all of the changes that people suggested last night (and this morning) and it's taking longer than I thought. I hope to have it ready in a couple of hours.--Dallan 14:00, 20 August 2009 (EDT)

Ok, the lists are finally updated today. You'll probably need to force a refresh of your browser to see them. They should include the new capitalization rules (new prepositions, except for the umlauts just mentioned, lowercase .com, .net, etc., leave acronyms uppercased, capitalize authors correctly), and the fix to the place issued for newspapers. I really hope to get them updated first thing in the morning from now on.--Dallan 17:01, 20 August 2009 (EDT)

Multiple authors of books [24 August 2009]

I have no problem with the source title only listing the first author; but since the bot is supposed to pick up the author from that field exactly where and how are we supposed to give credit to the additional authors? Also what did we decide on using compiler if it is compiled?--Beth 16:52, 20 August 2009 (EDT)

Please list all authors in the author field. The bot will make the first author part of the source page title. When we start generating full citations we'll list everyone found in the author field in the citation. For compilers, put the compiler's name in the author field followed by ", comp." or ", compiler". That way they'll be noted as a compiler in the full citation, but we'll use just their name (without the , comp.) for the source page title.--Dallan 17:42, 20 August 2009 (EDT)

Is the same true for editors? Can we put ", ed." or ", editor" after the name and have the system ignore it for the page title? -- Amy (Ajcrow)

I like the idea of ignoring the "compiler" or "editor" for the Page title. However, if that can't be done (or consensus /majority goes for including it), could at least be lower-case? Otherwise, it competes with the name of the compiler/editor/etc. See:

Source:Hill, Peter B., Compiler. Mayflower Families Through Five Generations

Thanks! jillaine 14:37, 23 August 2009 (EDT)

I just looked at your list of renames and it looks like the new title won't have the compiler in it. I haven't changed the code to make this happen for newly-added sources yet (I'll do it later this week), but words like compiler, editor, and translator should be removed from the title, as well as anything in parentheses at the end of the author field.--Dallan 12:37, 24 August 2009 (EDT)

School Records as a Subject [20 August 2009]

As I was editing Source:Admission registers of Wesleyan Boys' School, 1912-1941, I discovered that there is no option for school records in the Subject field. Could this be added? -- Amy (Ajcrow) 17:16, 20 August 2009 (EDT)

I could, but I've been using "Institutional records" for this type of record. Would that work ok?--Dallan 17:42, 20 August 2009 (EDT)

I've been doing that, too -- in my case, with records of masonic lodges. Any organized entity that is neither a government agency nor a church agency is an "institution," which is simply the more general term. --Mike (mksmith) 17:49, 20 August 2009 (EDT)

FHL microfilm roll numbers [20 August 2009]

I know I saw a discussion about this somewhere today, but I'll be darned if I can find it now. Are we supposed to removed the FHL microfilm roll numbers from the title? -- Amy (Ajcrow) 17:23, 20 August 2009 (EDT)

According to the proposed new title (the second, indented line) of those on my list, it appears the parenthetical numbers are being whacked off automatically. --Mike (mksmith) 17:51, 20 August 2009 (EDT)

That's right -- the renamed titles shouldn't have them.--Dallan 18:35, 20 August 2009 (EDT)

"Saint" vs. "St." in place names [20 August 2009]

I can't remember now what its previous behavior was, but the drop-down list of place-names is insisting on spelling out "Saint," contrary to ordinary and official usage. E.g., "Saint James Parish, Louisiana." The parish and state websites both use the form "St. James," and so does the St. James Parish letterhead (on physical paper). We also have St. Mary and St. Tammany, and others. --Mike (mksmith) 18:53, 20 August 2009 (EDT)

The drop-down list appears to not work with St. James, even though it is listed as an alternate name for Place:Saint James, Louisiana, United States. I'll have to figure out what's going on later. I haven't changed anything with the drop-down for quite awhile.

Since you say that St. James is the "official" name, I just renamed Place:Saint James, Louisiana, United States to Place:St. James, Louisiana, United States, although the towns in St. James as still titled using Saint; e.g., Place:Batree, Saint James, Louisiana, United States. They could be renamed as well if someone was really interested.

Length of a page title? [20 August 2009]

I assume there's been a length limit for page titles all along, but this is the first time I've bumped up against it:
Old form: Source:American State Papers: Documents, Legislative and Executive, of the Congress of the United States, in Relation to the Public Lands
New form: United States Congress. American State Papers: Documents, Legislative and Executive, of the Congress of the United States, in Relation to the Public [title truncated here]

I gnawed at this awhile when I first created it, because of the length. But I wanted to distinguish between this one (the land volumes only, with different pagination) and the very similarly named full set, which will be just "United States Congress. American State Papers." You have to get out into the subtitle before there's any significant difference in the two. If I were citing the land-only set on paper, I could replace of the Congress of the United States, with an ellipsis and be done with it, but that's too arbitrary to work here. I can't just add the date of publication, either, because they're not two versions of the same book. Is it going to upset the system if there are two closely related but different works with almost-identical titles? --Mike (mksmith) 19:13, 20 August 2009 (EDT)

It won't upset the system as long as there's some difference between the two titles.--Dallan 23:19, 20 August 2009 (EDT)

Dang it, I've been doing it wrong! [20 August 2009]

Gosh darn it. I misunderstood and I've been doing it wrong.

I thought we were to be cleaning the source pages up so that they'd be READY for the auto-renaming. Now I'm believing I should actually be DOING the renaming.

Please review what I've been doing and advise. DANG DANG DANG! I'm SO sorry if I've been messing it up. FRIG!!!!

jillaine 21:21, 20 August 2009 (EDT)

Before you hurt yourself (or the wall), calm down. I think you are ok. We were supposed to change the content fields so that the rules can make the title look correct on the next run - not do it yourself. I have had to do some renames on some of mine because they had other problems. For example, add a date for edition. --Judy (jlanoux) 21:41, 20 August 2009 (EDT)

(kahlua, cream and graham crackers later... yes graham crackers... i no longer feel the bruises on my forehead), so just explain this to me, please. On the page with our list, there's the link and immediately below it I assume is what the rename should look like. Is that rename what WE'RE supposed to do, or are we to get the page set up so that when the auto-rename runs, it will end up looking like that second line? Thanks. jillaine 22:38, 20 August 2009 (EDT)

The second, indented line is what the renaming robot is proposing that the new page title will be -- based on what it finds in the text boxes on the page. The robot is asking you, "Is this how it's supposed to look?" If it's correct according to the rules on the Project page, then you don't have to do anything. That was the case with most of mine, since I was the creator, and it will undoubtedly be the same with your list. Anything that looks odd is likely a glitch in the robot's programming -- all that stuff people are noting on the Talk page here. That stuff, bring to Dallan's attention. You don't try to retitle pages yourself because the robot has to deal with the identical problem, probably, on many other Source pages. We have to get the machinery to function properly, not just manually fix a single page. And when Dallan fixes the programming, the robot goes through it all again. Tomorrow morning, skim back down your list. The robot is now showing you the revised version of anything you changed and is asking, "Is it correct now?"

The fun part, actually, is when you finish with your own list and tackle one of the "Other" lists. Much slower. Most of those are turning out to need significant help. Almost all the sources types are "Miscellaneous," the publisher/place/date are all crammed in a single field, there's no punctuation in the title (so the robot can't see the obvious subtitle), and on and on. Lots of fun stuff. And any of those pages that you edit -- which I find is virtually all of them -- is going to find its way to your list tomorrow morning, since you were the last one to fiddle with it. Cheers! --Mike (mksmith) 23:13, 20 August 2009 (EDT)

<img align=center src="http://l.yimg.com/static.widgets.yahoo.com/shots/36981-shot.gif">
jillaine 23:32, 20 August 2009 (EDT)

Email - I'm so sorry [21 August 2009]

Don't worry about it - not a big deal. Things are moving around as the rules change and if we edit something it moves to our list. Everything has to be looked at again anyway since we keep changing things.

Splitting the lists was a good idea. Mike and I were working from either end toward the middle. I'm not sure how you did it, but the ones at the bottom that were already looked at stayed on list 0.

Before the next round (citations) can we have some lists of those where Genealogical Society whatever in Salt Lake has overwritten the publishing data. These will have to be fixed manually unless you can pull publishing info from some online catalog. We're focusing only on title right now. Fixing pub info takes a lot of time.

I can do that, but I think the list will be huge. The GSU is listed as the publisher quite often.

Well not ALL of them, but maybe the ones that are on the lists we're reviewing. If you could generate the personal lists (with the ones we're using) first and we get those fixed, we would probably have a lot fewer to deal with. It should be only the 'Book' type.--Judy (jlanoux) 12:00, 21 August 2009 (EDT)

In fact, it would be nice to have a list of sources cited on pages we watch so we could check everything of concern to us.--Judy (jlanoux) 07:20, 21 August 2009 (EDT)

That's a great idea. After the renaming is done I'll put together a list of sources cited on pages that people are watching.--Dallan 11:49, 21 August 2009 (EDT)

BTW, the email that Judy is referring to is an email that I sent to people who were reviewing the "Other" lists. I had broken up the 4 original lists into 8 smaller lists, but had neglected to update the links, so for a period of 6 hours people might have been reviewing the same list that someone else was also reviewing.--Dallan 11:49, 21 August 2009 (EDT)

World War I Draft Registration Cards [22 August 2009]

Can we come to an agreement about how to title these pages and fill in the fields?

Please see what I did (on the fields) here:

Source:Kansas, World War I Selective Service System draft registration cards, 1917-1918

By doing so I'm proposing the following:

• Type: Government/Church records
• Author: Selective Service System
• Title: World War I Draft Registration Cards, 1917-1918 (STATENAME)
• Place issued: Washington, DC
• Subject: Military records
• Place(s) covered: STATENAME, United States

The overall source for all WWI draft reg cards would be the same as above, less the state info

Right or not?

Thanks.

jillaine 08:35, 21 August 2009 (EDT)

Hi Jillaine, there is one already setup; correct or not I am not sure with all of the nuances of source page titles. Here is the name. Source:United States. World War I Draft Registration Cards, 1917-1918--Beth 09:35, 21 August 2009 (EDT)

The above source page has alot of information on it; so I would correct this one first and then title the others to link to this page. I removed the place name from the title, but I see you have a different title so I guess we need a consensus on the title, etc. Doesn't matter to me. --Beth 09:40, 21 August 2009 (EDT)

I didn't realize this broader page (with all the info) existed; I'm glad to know it's here. And since the draft registration was carried out to a close national standard, we probably need only this one page. State and county info would be "citation detail." And I would leave the author/title as it is for this one, even though it might not be technically correct, because "World War I Draft Registration" is what absolutely everyone calls it. --Mike (mksmith) 10:34, 21 August 2009 (EDT)

Okay, if nothing is pointing to a state-level WWI reg card page, I'll delete the page; if something is pointing to it, I'll redirect any state pages to the overall page. Thanks everyone for the quick resolution. jillaine 12:29, 21 August 2009 (EDT)

I think we're going to want to include this on the Help page for creating sources, to discourage people from creating local draft registration pages in the future. Maybe on the Source Portal page, too. --Mike (mksmith) 13:47, 21 August 2009 (EDT)

I agree (at least on Help:Source page titles). Would you mind adding it sometime?--Dallan 13:59, 21 August 2009 (EDT)

Regarding updates to tutorials, help pages, etc. requested by you Dallan from time to time, we need a location to cut and paste your requests so volunteers could work on the items and check them off when they have a moment of free time. --Beth 21:36, 21 August 2009 (EDT)

Great idea. I've created WeRelate:Help project. I'll start adding things to that page.--Dallan 01:01, 22 August 2009 (EDT)

What happens with duplicates? [2 September 2009]

There are multiple records for some sources and the renaming project should give them identical page titles. They aren't necessarily on the same list now. What is going to happen? Does it break something? --Judy (jlanoux) 08:38, 21 August 2009 (EDT)

Yes, is breaks something :-). We can't rename two pages to the same title. There are approximately 50,000 source pages that currently will have duplicate titles post-renaming. The vast majority of them are for FHLC titles that are government/church records with the same place-covered and title fields. I've looked at some of them and the only difference I can see is a small difference in the author. But they're generally in foreign languages so I can't tell whether that difference is indicative of different sub-jurisdictions of the place covered, or just a different person in charge of the agency. I've decided to "punt" on those and keep the FHLC number at the end of the title so we don't have to worry about them.

Some of the duplicates may have to be merged because they represent the same book or record collection. I thought we could look at these after we got the sources in the current lists reviewed. We could even start the renaming process on the non-duplicate sources while reviewing the duplicates. I'll try to post a list of duplicates that excludes the FHLC-only duplicates by the end of the day today.--Dallan 13:16, 21 August 2009 (EDT)

I just posted a list of duplicates . It's using data that's a few days out of date and I've already taken care of the duplicates on the first few screens. There are approximately 1900 pairs of duplicates in the list. Most of them are due to Ancestry and FHLC having the same source, so resolving the duplicate will involve copying the repository from one source over to the other source and deleting the first source. We don't need to do anything with it yet; I just posted it in case people were interested in looking it over. I'll post an updated list with up-to-date information tomorrow morning.--Dallan 22:30, 21 August 2009 (EDT)

There may be a problem here. There's no way many of those census Source pages are duplicates -- Source:United States, Indiana, Greene. 1850 U.S. Census Population Schedule and Source:United States, Tennessee, Scott. 1850 U.S. Census Population Schedule, for instance. The authored books, yes. --Mike (mksmith) 10:48, 22 August 2009 (EDT)

The problem here is that those two sources don't have their source type set, so the new page title is getting constructed from the title field only. If you set the source type on these pages (to g/c records) that will fix it.--Dallan 13:32, 22 August 2009 (EDT)

Dallan or Solveig, could you re-run the Duplicates list, please? I see that Judy has resigned from this project for now, and I'm willing to step in (or should I work on ceyockie's list first?) since I've finished my other source review pages. jillaine 10:43, 31 August 2009 (EDT)

I've just re-run the duplicates list with the census duplicates removed. There are roughly 600 sets of duplicates left. The list is in alphabetical order; perhaps you and Amelia (and others?) could divide up the list by alphabetical range. Now that we're back, Solveig can help out as well.

Thanks!--Dallan 17:55, 31 August 2009 (EDT)

I've just updated the main project page with an area to sign up for resolving duplicates for a specific range of the alphabet. I would really appreciate people signing up to resolve these duplicates. As I see it, this is the last big review task left. Thank you!--Dallan 18:32, 31 August 2009 (EDT)

Can I just say that this is one doozie of a project-- deduping these sources. I admire Judy for her effort, understand why she'd resign from it, and feel humbled about how slowly this goes. I've got a rhythm down, but it's still slow. jillaine 09:54, 2 September 2009 (EDT)

Amazing, isn't it? After doing nothing but this project for a week, I had to quit so we could get packed up for our trip.I went through the top half of the alphabet about three times and every day more would appear. I'm still looking at all those Russian ones. I think I'm going to consolidate them into a single source as the only difference is the years-its really one recordset. But probably won't be able to get to it until the weekend. --Judy (jlanoux) 16:32, 2 September 2009 (EDT)

Judy, don't bother with those Russian ones in the A's. Amelia already communicated with me that she's going to take care of it (doing as you've described, I believe). Enjoy your trip. -- jillaine 21:49, 2 September 2009 (EDT)

Mysterious Comma in new source page title [21 August 2009]

See this on my project list.Source:United States, South Carolina, Barnwell. 1840 U.S. census, population schedule Note that I edited the title, which does not include a comma, but the newly generated source page title includes a comma. What comma monster has inserted itself into my edit?--Beth 09:25, 21 August 2009 (EDT)

And rendered the title in lowercase, too. Weird. --Mike (mksmith) 10:36, 21 August 2009 (EDT)

I just looked at your list, and the new title will be "Barnwell, South Carolina, United States. 1840 U. S. Census Population Schedule". The comma has been removed. Based upon when you wrote your message, I think the problem was the list you were looking at hadn't been updated yet. I updated the lists today about an hour and a half after you sent your message. From now on, the lists should get updated automatically about 6am (EDT) every morning.--Dallan 11:55, 21 August 2009 (EDT)

Census (Basic Question) [22 August 2009]

This raised a question for me... a possible "oops." I have always thought the correct census title was YEAR U.S. Census Population Schedule. Then when I started this project and looked at one of Dallan's census records for a model, he had the title as YEAR U.S. Federal Census Population Schedule. So I started thinking maybe I should be adding the word Federal (maybe the rules had been updated when I wasn't looking??). But now I see above just U. S. Census Population Schedule.

Does it matter? Is U.S. one word (no space) or two? [U. S.] Does it matter whether we're consistent always? Thanks...Brenda--kennebec1 13:01, 21 August 2009 (EDT)

Oh please don't use my source as an example. According to Help:Source page titles, the title should be:

XXXX U.S. Census Population Schedule

Consistency isn't critical, but it's nice.--Dallan 13:27, 21 August 2009 (EDT)

What is the format for Canadian censuses? -- Amy (Ajcrow) 14:00, 22 August 2009 (EDT)

If there isn't a Canadian census format on Help:Source page titles (I'm typing from my phone and it's not easy for me to check) then could you define one along the same lines as the other census formats?--Dallan 23:16, 22 August 2009 (EDT)

Lists updated [21 August 2009]

I've updated the lists this morning, and updated the links to them as well.

The lists have been moved to the main server, where they will be refreshed automatically every morning from now on at 6am Eastern time.--Dallan 10:51, 21 August 2009 (EDT)

Serial Question (Books), etc [21 August 2009]

Hello All,

I am using Books for Sources that are multi book index's, instead of the misc. If I understand that is what is needed or preferred?

Do you mean a later-written index to an old work that is itself published as a book? If so, the index is a "book." --Mike (mksmith) 12:52, 21 August 2009 (EDT)

Hello & Good Morning Mike,

Yes. An example is the Index to the DAR Lineages Volumes. The Index's are usually a three volume book index. Which indexes the many, many volumes of DAR Lineages. --DFree 14:21, 21 August 2009 (EDT)

Ah. Yes, that's a single (VERY large) book. All the volumes of lineages + the 3-vol index get only a single title and a single Source page. --Mike (mksmith) 18:10, 21 August 2009 (EDT)

Also I might be screwing up because I am renaming (English language only) the web page (top rename) to get rid of the FHL #. It is not automatically going away. Should I stop that? --DFree 14:21, 21 August 2009 (EDT)

Have you looked down in the title field? I've had a number of pages where the number was in parentheses following the text in the title box -- which is why it isn't going away. Renaming the actual title won't have any effect since the robot is going to do it all over again anyway. --Mike (mksmith) 18:10, 21 August 2009 (EDT)

We are only editing (usually) the right-sided fill-in-the-boxes content? Title, subtitle, author, etc. Correct?--DFree 14:21, 21 August 2009 (EDT)

We do not add . (periods, etc) because the computer will do that Yes? --DFree 14:21, 21 August 2009 (EDT)

Based on the way the system has been constructing census pages, I think periods are being added automatically. But I hope there's also a line of code in there that will check for duplicated punctuation marks, just in case. --Mike (mksmith) 12:52, 21 August 2009 (EDT)

Thanks for the feedback, still a little confused, Debbie Freeman --DFree 10:55, 21 August 2009 (EDT)

• Could you give an example of a multi-book index so that we could look at it?

Hello Dallan, See above to my message to Mike. The FHL catalog has the three volumes index as one listing. --DFree 14:21, 21 August 2009 (EDT)

Good point. For now, list it as a book. When we start generating source citations I'll have to add a field to the source page that lets you specify the number of volumes.

• You don't need to rename any pages. (It won't do any good anyway because the system's going to rename them all later.) The system should be getting rid of the FHL# when it does its rename.

-- You are renaming later automatically Correct? --DFree 14:21, 21 August 2009 (EDT)

Yes. The text underneath the source link in the list is the title the source will receive unless you edit the fields on the page. The lists are updated each morning with new titles based upon the edits you did the previous day.

• Yes, you're editing the right-side fill-in boxes. The lists show the current source page title (the link in blue) and what the new source page title will be after the renaming (the text underneath). In many cases the new source page title isn't very good, either because the source type has been set incorrectly, or because the author, title, and/or place covered fields are empty or incorrect. Your job is to edit those fields so that when the system renames the source according to the new source page title rules about a week from now, the source will get the correct page title.

Thanks for the clear instructions. I appreciate it. In a case were no author is known we leave blank? Or add unknown? --DFree 14:21, 21 August 2009 (EDT)

I'd leave it blank.

• The computer will add the periods between author or place-covered and title. And it has a check for duplicate periods. (If you notice duplicate periods in the new source page titles displayed in the list, please let me know.)

Thanks!--Dallan 13:52, 21 August 2009 (EDT)

I can see later I will need to go back and add a more correct WeRelate Source to some of my Person and Family Pages. An example is the 1832 Sussex England Tax Source. I need to create a new Source which is the title of the book which the tax list is in. --DFree 14:21, 21 August 2009 (EDT)

See the above (long) discussion about Government/Church records. If the book was authored by a government agency and not a human being, you should leave it as a government/church record type and just add the author (and publisher) of the book.--Dallan 15:11, 21 August 2009 (EDT)

-- Hello Dallan,

My theory is that this 1832 Sussex, England Tax List source exists. So I would keep this source. But add the authored book as the other source. I would then cite the authored book. The authored book is not a government book, more like our American County history book. Thanks for the input --DFree 18:18, 21 August 2009 (EDT)

Fixing titles that start with United States [21 August 2009]

Just an FYI, User:Moonknight125 (one of my children) is reviewing the 700 sources whose titles begin with "United States" followed by a period or a comma, and removing the place from the beginning of the title if it makes sense. He'll hopefully be done by the end of the day today.--Dallan 13:41, 21 August 2009 (EDT)

Sounds like Taylor going off to college means all your other kids got promoted. . . . --Mike (mksmith) 13:51, 21 August 2009 (EDT)

You got it :-). Today is Taylor's last day at home. We leave tomorrow to drive him to college. And User:Skater has been missing all the fun around here because she flew out there a couple of weeks ago to spend some time with friends.--Dallan 13:57, 21 August 2009 (EDT)

Fish or Fowl? [21 August 2009]

Okay, guys, the two hemispheres of my brain are getting into a shouting match over Source:Descendants of Edward Ball of New Jersey Interest Group. The Big gEDcom. Is this a "source"? Or a "repository"? Or something else altogether? Nothing links to it except a redirect, and I'm half tempted to kill it off and bury the body when no one's looking. --Mike (mksmith) 13:57, 21 August 2009 (EDT)

I thought single family gedcom sites like this were supposed to be deleted when we did the last source review, but it looks like this one had a human author, so perhaps that wasn't done. I vote to chuck it. If someone wants to cite it, they can resurrect it themselves.--Amelia 14:01, 21 August 2009 (EDT)

They were supposed to be deleted. In cases like this where it doesn't qualify as a community Source page, if nobody links to it, then delete it. Otherwise change the type to "Should be a MySource" and we'll contact the person linking to it when the rename is finished.--Dallan 14:08, 21 August 2009 (EDT)

If that's what this thing is, then it's gone. --Mike (mksmith) 14:39, 21 August 2009 (EDT)

What's going to happen to the other ones? [22 August 2009]

This is a little off topic, but what's going to happen to non-human-edited sources that don't fit the format now? For example, Source:Citizens of Milton County, Georgia 1860 census index We have a format for census records, but this is currently listed as "miscellaneous" with the wrong county because it changed names later, and the transcriber as an author. Are we going to fix these by hand, or will the system fix the census records to the correct format? And if it's not, then at least my time could be better spent working on those fields than the ones on my own list that might have a stray character or two, but at least will get the right place and a close title. --Amelia 15:27, 21 August 2009 (EDT)

Except this isn't a collection of census records. It's an index/finding aid, an "after-market" publication. Like all the AIS indexes we used to rely on before Ancestry indexed the census. I would treat it like any other book. --Mike (mksmith) 18:21, 21 August 2009 (EDT)

Mike, I have to correct this. I know a librarian makes this a separate source, but that's not the rule here, and it simply can't be. We can't have a separate census source page for every transcription out there. That would add thousands of records to the catalog, while losing the benefit of having the transcriptions linked to the actual census source and the common naming and category hierarchy that puts all the censuses together. An authored transcription is never going to be found or cited except by the most exacting researcher, and the chances of someone citing a non-circulating book at the FHL is even smaller. I've never seen it in dealing with hundreds of these records and person pages citing the census. The community is far better served by renaming these as census records with the FHL book, or whatever, as one repository. (We could delete them as useless--indexes don't get their own source pages--and recreate new ones from scratch, too, but that seems like a bit of a waste).--Amelia 11:03, 22 August 2009 (EDT)

There are roughly 900,000 non-human-edited sources to rename. The system is going to guess whether to use the author. title format or the place. title format for these sources, based upon whether the author looks more like a person/genealogical society or a government/church agency. You can see a sample of 9,000 sources here. If you're interested, would you mind reviewing the sample to see if you notice any general problems with how the system is guessing? I could post a larger sample if you're interested.--Dallan 01:22, 22 August 2009 (EDT)

It's a systemic problem not to have census records and other standard sources not renamed to the agreed-upon format, but I think you're saying that won't be fixed by the system. And actually, they're a pretty small sample of the ones you posted, so renaming them as they come along is probably not a huge deal.--Amelia 11:03, 22 August 2009 (EDT)

Yes, I think we'll have to live with standardizing the titles manually (but not now).--Dallan 13:36, 22 August 2009 (EDT)

Authors on a Government/Church Records page [21 August 2009]

If there is an author on a Government/Church Records page, will it trump the geographic designation? In other words, will the page title end up as Place. Title or as Author. Title ? -- Amy (Ajcrow) 15:44, 21 August 2009 (EDT)

If the designation is "Government/Church Records," the renaming robot will take the geographical info and the title. It will completely ignore whatever might be in the author field. (I checked on that earlier myself.) But if there is an author, you should look more closely at the rest of the info on the page to make sure it's not really a book that presents a collection of transcribed government or church records. I've had a number of those, too. --Mike (mksmith) 18:25, 21 August 2009 (EDT)

Periodicals [23 August 2009]

I know that recently we discussed Source Pages for periodicals vs Source Pages for articles. When we have a Source Page for a periodical, how do we title it? Help:Source page titles says to enter the name of the periodical (spelled out). That works fine for something like the New England Historical and Genealogical Register or the National Genealogical Society Quarterly. However, there are numerous periodicals that don't contain the name of the society. The problem is that there are periodicals out there with the same title. How can we differentiate the Source Pages of Family Branches published by ABC Society and Family Branches published by XYZ Society? -- Amy (Ajcrow) 16:56, 21 August 2009 (EDT)

On the first question, I have my own example: Source:Lifeliner (Genealogical Society of Riverside, California). It presently includes the name and place of the society, but the renamed version will be just "Lifeliner." But I think it's basically an all-or-none proposition. You would have to set the bot to add location or society name info to all periodicals -- assuming they were designated "Periodical" on the page. But the search-results list includes other information than just the title, so the user/searcher can decide which one they want. On the question of duplicated titles, though, I'm pretty sure that's a no-no. A certain sub-set of periodical titles -- those the bot flags as dups -- may have to be treated as if they were newspapers, by adding the place after the title in order to distinguish the two. As in "Lifeliner (Riverside, California)." --Mike (mksmith) 18:36, 21 August 2009 (EDT)

The Lifeliner source looks like it will be renamed using the author-title format: "Genealogical Society of Riverside. Lifeliner (Genealogical Society of Riverside, California)". Since its source type is Miscellaneous the system guesses whether to use author-title or place-title for the renamed page title, and since it's authored by a society, the system uses the author-title format. If its source type were changed to Periodical, then the renamed page title would be include the title field only, which would be the same as the current page title.--Dallan 10:00, 22 August 2009 (EDT)

That poses an interesting question. Do we leave it as Miscellaneous and have it pick up the society's name or do we make it Periodical, which would leave off the society name and result in duplicate pages? -- Amy (Ajcrow) 13:37, 22 August 2009 (EDT)

For the sake of clarity and consistency, I think it would be better to differentiate the duplicates with the society name, as some societies cover more than one county (in which case, which one would you list?) or aren't even geographically based (thinking about ethnic societies). -- Amy (Ajcrow) 08:13, 22 August 2009 (EDT)

I agree. If I misinterpreted the way the stated rules will operate for society-published journals, I apologize -- and I'm happy to be corrected, because I would like to see those publications differentiated automatically by including the society name. --Mike (mksmith) 11:00, 22 August 2009 (EDT)

This is true for other items -- like City Directories -- as well. Different publishers publish directories for the same location (and even in the same year). In one example I created, I used the publisher's name as part of the title Source:United States, New York, Kings, Brooklyn. Lain and Co City Directory. Using a similar theory, one would put the name of the group (the publisher) in the title to distinguish an otherwise duplicate name.

What category should a city directory be? Is it a periodical? (i.e. an annual/semi-annual publication?) It could be a book, but it would have multiple editions (one for every year that publisher produced the directory). I have the above source set as Misc, but that leaves the source without a geographic name, and that is essential for a Directory. Perhaps there should be a Directory type? There is a Directory records *subject,* but I'm not sure how to categorize its type. -Brenda--kennebec1 18:28, 21 August 2009 (EDT)

A librarian would call a city directory a "serial," which is a category of (usually) subscription items that also includes almanacs, yearbooks, city ordnances, and other items that are issued "continually." What we commonly call "periodicals" are just a sub-category of serials. A serial gets cataloged as as single title, like a single book. However, we don't have that elaborate a distinction. (And it really only matters to librarians.) For our purposes, I think I would call a city directory a "book," and then list the first and last years of publication for a given town in the text box (or something like "1898-     ." if it's still being published). You certainly don't want a separate Source page for each year's volume. And the annual issues aren't the same as "new editions," either. Also, most city directories seem to have distinctive titles, including date, on the title page -- "West's City Directory of Zanesville, Ohio, 1900" -- so keeping the various Source pages separate ought not to be a problem. I don't think the "directory" subject is the same thing as a city directory, actually; in fact, I'm not sure what it's supposed to encompass. --Mike (mksmith) 19:24, 21 August 2009 (EDT)

Why wouldn't a city directory be a directory? -- Amy (Ajcrow) 08:13, 22 August 2009 (EDT)

I think the question here is what source-type to use. I assume the subject would be directory.--Dallan 10:00, 22 August 2009 (EDT)

Okay -- but what exactly does "Directory Records" mean? (A question for some later time, actually.) --Mike (mksmith) 11:04, 22 August 2009 (EDT)

Currently the system titles pages for periodiclss using the source title only. Should I change it to include author (like books)?

That sounds good. It would help avoid duplicate pages and it will make it easier for people to create sources later. -- Amy (Ajcrow) 13:57, 22 August 2009 (EDT)

Ok, I'll add author to the page title for periodicals. It will start showing up on Monday.--Dallan 23:21, 22 August 2009 (EDT)

I added the directory subject for lists of people that didn't seem to fit in other subjects, like city directoriezs. But I'm ok with merging it into another suject as well.--Dallan 13:48, 22 August 2009 (EDT)

I'm adding a vote for including not only the Author field for Periodicals, but all the fields a Book source has. I've got this:

Source:Colonial families of America (currently "Misc" because of all the fields I need).

that has an editor and a series of years it was published. But as the form currently stands, I can't add either.

Thanks!

-- Jillaine jillaine 23:19, 22 August 2009 (EDT)

I'll do this tomorrow evening.--Dallan 02:46, 23 August 2009 (EDT)

Edition year - where does it go [22 August 2009]

I'm trying to eliminate some duplicate books. I think this is where many of our potential duplicates are - I've seen six copies of one book! I'm confused about how to get the (year) added to the title without losing it in the rename. Renaming the page didn't work. I have settled on adding it to the end of the title in the database box. Is that correct? --Judy (jlanoux) 17:43, 21 August 2009 (EDT)

I understood that we were supposed to end up with one source for a title, and somehow reflect in the description the various editions. I've been going with the earliest (known to me) edition as the "master" edition for the right side form, and adding info on later editions in the text. Any thoughts on how this could be done more efficiently? - Brenda--kennebec1 18:10, 21 August 2009 (EDT)

There's a difference between "editions" that are substantively different and ones that are just reprints. The former get separate source pages, which is where the year comes in. You do put information on later editions in the main text box (ideally), linking the editions together. Reprints all go on the same page. And, yes, Judy, you're correct. You have to add the year to the title box.--Amelia 18:14, 21 August 2009 (EDT)

Er, well, I thought that was the case, but then I realized that one source I recently added that needs a year isn't on my "non-matching" human review list, which means, I think, that the rename wouldn't change the name. And it doesn't have the year in the title field. So someone else needs to answer Judy... --Amelia 18:18, 21 August 2009 (EDT)

Brenda, you have to understand the arcane world of books or spend a lot of time on catalog web sites (in my case) to figure out which is correct. If you look at the current featured page, there is one edition which was reprinted many times in various formats as a photoreproduction (pictures of the original pages). So we only use one entry. In contrast, one I just worked on is Source:Genealogies of the families and descendants of the early settlers of Watertown, Massachusetts and Watertown records which was issued in 1855 and a 2nd edition was issued in 1860. Since then NEHGS has reprinted the second edition several times again via photoreproduction - all of these live on the page with the 2nd edition. So this source has two pages. One for the 1855 edition and one for the 1860 edition. These old books can get really confusing because they are reprinted many times and someone may add new material and create a new edition along the way. I use Library of Congress catalog, Amazon, Google among others to look up books I am not personally familiar with.

There are a few librarians here so don't hesitate to ask if you need help deciphering something. Sometimes the genealogy of the book itself is more interesting than the contents.--Judy (jlanoux) 18:38, 21 August 2009 (EDT)

Part of the renaming code looks for a four-digit year in parentheses; e.g., (1931) at the end of the current source page title and if it finds one, it keeps it in the renamed title. So you could either add the year to the end of the title field, or rename the page to include the year in parentheses at the end of the page title.--Dallan 09:51, 22 August 2009 (EDT)

Vote on Massachusetts Vital Records Format [23 August 2009]

For consistency, please vote (or tell me the rule) on a format for the Tan Books:

1. Dartmouth, Bristol, Massachusetts, United States. Vital Records of Dartmouth, Massachusetts, to the Year 1850
2. Dartmouth, Bristol, Massachusetts, United States. Vital Records of Dartmouth, Massachusetts to the Year 1850
3. Dartmouth, Bristol, Massachusetts, United States. Vital Records, to the Year 1850
4. Dartmouth, Bristol, Massachusetts, United States. Vital Records to the Year 1850
5. Dartmouth, Bristol, Massachusetts, United States. Vital Records to 1850

(All those combinations of commas, shortening, and not are found in the current titles; I'm not promising to rename them all, but it would be nice to get consistent on the one's we do. And, if you're so inclined, I started Category:Massachusetts town vital records ("Tan Books") while at this.)

Also, for anyone following this area, note that the "vital records to 1850" books by Holbrook will not follow the geographic format and will be under his name.--Amelia 18:33, 21 August 2009 (EDT)

Ok, I'm thoroughly confused again. Not being a New England researcher, I have no idea what a "Tan Book" is. Can someone please list how we are supposed to treat the various types of New England "Vital Records of XXXX" (whether "to the Year 1850" or not)? Thanks! -- Amy (Ajcrow) 08:24, 22 August 2009 (EDT)

The Tan Books are the series of Massachusetts Vital Records described here. It is a question of standardization because the titles are supposed to be the same, but the FHC's use of commas and our own rule about dropping the place in the title create some opportunities for inconsistencies. Other vital records, to my knowledge, don't have the same problem. They should be titled as according to the rules for either gov/church or book as appropriate.--Amelia 00:12, 23 August 2009 (EDT)

I vote for #5; short and sweet. jillaine 06:52, 23 August 2009 (EDT)

I'd say #4, tho 5 would be ok. I think its helpful to have the name be more closely parallel to the original (minus the place name), as that helps the person looking on a dropdown list be certain they have found the correct source. - Brenda --kennebec1 08:09, 23 August 2009 (EDT)

I would not block consensus for #4 (she writes while waiting for Quaker mtg to begin)-- no I'm not inside but out enjoying a brief weather respite. jillaine 09:24, 23 August 2009 (EDT)

Back to Braintree compilation of vital records - Clarification? [22 August 2009]

Ok, I read over all of the government/church record discussion and I *think* the consensus is that my particular example Source:United States, Massachusetts, Norfolk, Braintree. Vital Records and Church Records of Braintree, Massachusetts is NOT a government/church record, because it was compiled by an individual from a variety of sources rather being an official record.

Thus Waldo Sprague's records would be a source type of "book" or a manuscript. Presumably a manuscript because it is unpublished? That's still a bit unclear to me.

But in this case, the NEHGS is not providing images of Waldo Sprague's original compiled records/manuscript. Instead they are providing an extracted database created from Waldo Sprague's notes (one more generation removed from the original).

So, I think this makes the Vital Records and Church Records of Braintree, Massachusetts a website. Maybe there will be another source someday from the original records at NEHGS, but in any case, I've made clear what the background is of the database on the source page. Does this make sense?

thanks, -Brenda --kennebec1 19:45, 21 August 2009 (EDT)

While I agree with your analysis, in a vacuum, I don't think making this a website serves any of the purposes that the page titles are supposed to serve. We use "website" when we can't think of anything better, as far as I can tell. Websites that cover a geographic area get the geographic records format of "gov/church records". Not all "gov/church records", despite the name, are "official" government sources; they include lots of records -- like most of the Massachusetts vital records -- that were privately compiled from various sources. If you want to go with Author. Title using Sprague, that too would be findable, and this version of Sprague would be a repository in addition to the original. We can't have separate source pages every time a website makes a transcription, it would get nuts quickly.--Amelia 20:10, 21 August 2009 (EDT)

I agree with Brenda's analysis. The published Mass. VR are not government records, they are books. What you get depends somewhat on the compiler, like the decision not to include the actual marriage intentions, but just to star those marriages that do or don't have matching intentions somewhere in the town records. And the published ones contain all sorts of additions, like all those PR (personal record) entries, that make them something else than the government records. Sometimes the entries are reworded to fit the entry to a pattern that prints nicely on one line.

Likewise, the Braintree collection, which I am unfamiliar with, is clearly the work of an individual, not the government, so clearly not a government record. But since Brenda is looking at a transcription of it on a website, I think she is exactly right to flag it as a website. Having spent some time comparing NEHGS's database of Vital Record of Rhode Island with the book, the format alone guarantees that it is not quite the same. Splitting it into records to fit it into a database tends to lose some of the context at least, in this example, taking a single record containing both birth and death, and creating one birth record and a separate death record, neither of which is a faithful reproduction of what is in the published book.

I think if it is published then it is a book. If it is a transcript on a website, it is a website. If it is the collection of an individual but unpublished, a manuscript. Only the original records, or photographic reproductions should be flagged as government records in my opinion. I think it would not be unreasonable to argue that even when a published version is commissioned by the same government agency, it is still a book, since some amount of editing/compiling is implied. For example, those records by the Record Commissioners of Boston. Somebody had to convert the cursive writing into type and format it to fit the media. This introduces room for error. Although I admit at least in this case there is some amount of official certification implied... --Jrich 21:50, 21 August 2009 (EDT)

I vote for website in this case. I've had to do the same for several I've found for HQ or Ancestry where they have put several things together and created a new database and named it. We have to go with their name. Otherwise no one can find it. --Judy (jlanoux) 09:25, 22 August 2009 (EDT)

Please use website as sparingly as possible. Websites aren't like books; they change their titles frequently, making it very difficult to find website sources a few years from now (so people will create duplicates). Making this source a "Government/Church record" adds the place-covered to the beginning of the title, which will make it more easily findable. I could add "Government/Church records - Website" as a new source type and have it result in the same page title as "Government/Church records": with the place covered at the beginning of the title. Or we could agree that the source type of "Government/Church records" includes website transcriptions as well as official transcriptions. Either of these alternatives is preferable to calling something a website and having its page title be based upon the website's title only.--Dallan 09:45, 22 August 2009 (EDT)

If I don't hear any objections, I'll add a new "Government/Church records - Website" source type tomorrow for unofficial (would like a better word here) transcriptions of government/church records on websites. We'll include the place-covered in the source page title for these websites. Sound ok?--Dallan 12:18, 22 August 2009 (EDT)

Sounds like a perfect compromise. --Judy (jlanoux) 12:55, 22 August 2009 (EDT)

I agree with Judy. This sounds like a very good solution all the way around. -- Amy (Ajcrow) 13:24, 22 August 2009 (EDT)

Your solution is fine, Dallan, but I'm not sure it addresses the underlying conflict we are having over the definition of Government/Church records. Some people see that label as applying ONLY to "official records." (the Tan Books, the actual images of vital records rolls/books from towns available on NEHGS, etc., the church publishing early records). Others follow a broader analysis - calling any collection, transcription, or database of official records (even when combined with private records), such as in the plethora of Vital records of _(insert place name here)_ books created by historical societies, albeit often with government sanction, as well as websites like ma-vitalrecords.org, or NEHGS, which post those books.

For genealogical purposes, there both is -- and is not -- much of a difference. It seems clear that access to images of official records, or the actual records directly, is the only authoritative (original) source (caveat: given the varying probability of misreading old handwriting, or local transcription and spelling errors or later editing, etc.) Everything else is at least one step removed from the source. Some of these items "one step removed" are books, others are manuscript collections at NEHGS or other repositories, others are compiled databases on websites, or transcriptions (of books or of original records) on web sites.

I think it's not so much that we need a Government/Church records - Website type, but that we may need a "Government/Church records - "Extracted" or "Secondary" or "Database" type to clarify this distinction. Or maybe we need a "Government/Church records - "Originals" or "Primary" type to distinguish official records from unofficial records.

Of course, I suspect that most of our government/church records sources are actually of the secondary (one or two degrees removed from the original, like my Braintree database) type. All of the books published by historical societies (Vital records of (place) to 1850 or 1892 or year XXXX) would be of this secondary type. But it makes sense to categorize these types of government/church records in some way that ties them to the same kind of naming structure (Geography. Title), because they function the same way as the "official" or primary records (which there are fewer sources for).

Sometimes (perhaps often) these secondary records are on a website like NEHGS or Ancestry as a database, but I'm not sure that these databases are that different from the historical society books in functionality, although they may be one more step removed from the original (the comments above regarding the NEHGS database based on the Vital Records of Rhode Island book is a good example of the problems caused everytime you add one more level from the original...).

But I don't think it's very user-friendly to ask someone to analyze their sources "degree of separation from the original" before finding a source. A user should be able to find anything that is vital records/government/church records/census records related in a similar format: the (Geography. Title) format.

All of which is a longwinded way to say, call it what you will, but ultimately I suspect we will need some sort of distinction between primary and secondary and maybe even tertiary "government/church records." In fact, the best solution may be to keep ALL of these as type: Government/Church records, and add a field that allows someone who sets up/edits the source to describe whether this particular "government/church record" source is primary, indexed/abstracted, compiled from primary sources, but thus secondary, etc. This is similar to the field set up when you actually *use* a source on a person page, where you can choose to describe the quality of a source. (Although when I use that field, I keep looking for an option in between "secondary" and "questionable" - something like "indexed/database," indicating that the source is not the complete original record, but an interpreted index to those records...)

but in the meantime, until more decisions are made, "government/church records-website" will certainly do... Thank you, Dallan. - Brenda --kennebec1 15:34, 22 August 2009 (EDT)

Suggestion for word to replace unofficial (Dallan requested a different word); derivatives --Beth 15:41, 22 August 2009 (EDT)

I agree with Brenda that all of the records should be kept as Gov/Church, which we define to include all derivatives. As far as I recall, "Gov/Church Records" was shorthand for the need to cover records covering a geographic area, not "records created by a government or church". Given that, I don't like having a separate type that sets up a distinction that isn't always appropriate. If we're going to define the difference, have people not read it and pick the wrong type, why don't we just define the current type to include those records, have people not read it and pick the wrong type? [;-)] And I direct the secondary/primary discussion back to the rule that what's supposed to be the same gets the same source PAGE (with distinctions between the technically separate sources explained therein). --Amelia 16:10, 22 August 2009 (EDT)

____ Nice analysis Brenda. All these problems come about because people took that one "type" to imply contents of the record and not the type of media. When definition of terms isn't clear it is virtually impossible to get agreement. So we just need a compromise everyone can live with and get back to work. Re Dallan's proposal, I would agree with Amelia that maybe a different type is overkill. Can we perhaps put the simpler alternative back on the table? Adding (Website) to the end of the name would clarify the nature of the source. It isn't useless clutter if it is essential for identification. It seems simpler to implement (we could do it now) rather than creating another type option. The concept of consolidating the source records will erase this problem for good if we go that way. --Judy (jlanoux) 16:39, 22 August 2009 (EDT)

Have to stay exactly with the FHL title, or not? [22 August 2009]

Hello All,

I am wondering if I am creating a problem. The FHL Catalog titles on occasion are not the best (in my opinion). Since some of the WeRelate Sources were taken from that catalog. I am doing small edits, like adding the States names in the WeRelate title. Example is in the catalog it says State Society, I have been adding the name of the State - California State. Should I not do that? Will it confuse the automatic computer process? The edits make it clearer though. Thanks for the feedback Debbie Freeman --DFree 23:18, 21 August 2009 (EDT)

I don't change the titles of books, but I agree that some of the FHLC titles are poor. I you think that a title needs clarification, I think it would be preferable to add things to the end of the title than to the beginning. That way if people know the title and start typing it the source will still show up in the drop-down list. And I'd add things to the title as an exception, not the rule.--Dallan 01:39, 22 August 2009 (EDT)

"Computer printouts"? [23 August 2009]

Can we do something better than this?

Downe Township, Cumberland, New Jersey computer printout; births or christenings, 1848-1867 

converted to:

The Church of Jesus Christ of Latter-Day Saints. Genealogical Department. Downe Township, Cumberland, New Jersey Computer Printout

There are a bunch of these on the agent list. They are all vital records that mirror some other source, and at the very least should be under the vital records format it would seem -- it seems ridiculous that something called a "computer printout" is its own source - it's a printout of something else. At the very least, why did the births/christenings get cut off? Can they be fixed without human intervention?--Amelia 01:06, 22 August 2009 (EDT)

These should generally be fixed tomorrow. I noticed a number of these showing up in the duplicates list. The new process that will run tomorrow morning will not list various forms of the GSU as an author in the page title (but they list themselves in several different ways, so if you notice some forms still in there please let me know).

If a source title is short and the subtitle is also short, I'll add the subtitle to the page title. I may not get to that first thing in the morning, but if not I'll try to get to it on Sunday.--Dallan 01:33, 22 August 2009 (EDT)

I believe that I've fixed this now, but some titles in the duplicates list are so long that the title still gets truncated before the important material after the words "computer printout". I think we may have to shorten those titles by hand.--Dallan 09:23, 22 August 2009 (EDT)

I ran into these computer printouts last night. I've been deleting them as duplicates. I thought they were just printed database dumps and we already have the database as a source. Same data = same source? I see it as the same issue as a book on Ancestry, only we're going backwards in technology. These have probably been replaced now that the databases are universally available. --Judy (jlanoux) 09:19, 22 August 2009 (EDT)

Deleting them as duplicates seems reasonable. I could delete all sources with the words "computer printout" in them, but I'm not sure if we want to go that far?--Dallan 09:25, 22 August 2009 (EDT)

I can do them. I agree you don't want to let a program run amok. The duplicate list is fascinating. When I get through reading the riot at the opera, I'll get back to work. Thanks for running it before your long day. --Judy (jlanoux) 09:46, 22 August 2009 (EDT)

Hold on -- Are you adding the link to the appropriate source page before you delete? Because otherwise you're destroying the link between the "computer printout" source and the source the FHL used -- it's usually not the same title, but the FHL catalog entry usually tells you what it was. Otherwise, they may be obsolete, but they're not dups, and a lot of them are in the IGI, so they are likely to be cited (the Hampden-in-Arden cite is mine for this reason) - and that may be a reason to leave them as well, with the titles not being consistent and all--Amelia 11:36, 23 August 2009 (EDT)

Driving tomorrow [23 August 2009]

I'll be on the road most of the day tomorrow (driving 900 miles). I'll be online for about an hour in the morning but then I probably won't be online again until Sunday morning. And I really want to thank everyone. This has been a terrific project. It will be wonderful to see it complete!

--Dallan 01:47, 22 August 2009 (EDT)

I'm thinking that by the time you read through all that was written while you were driving, that you'd wished you'd just kept going....

-- jillaine 21:29, 23 August 2009 (EDT)

This morning's list not catching "FHL parentheses" at end of title [22 August 2009]

I've seen a couple of cases this morning in my refreshed list that the auto-renamer is no longer removing the FHL # in parentheses at the end of the Page Title (it was yesterday). Example:

Current: Publishments, marriages, births and deaths from the earlier records of Gorham, Maine (294666)

Proposed: King, Marquis F. Publishments, Marriages, Births and Deaths from the Earlier Records of Gorham, Maine (294666)

Whas'up?

jillaine 08:08, 22 August 2009 (EDT)

Many (thousands) of the FHLC sources have the same place-covered and title, so I'm putting the FHL# back into the title in order to distinguish between them when there are several FHLC sources that would otherwise have the same title. However, I just realized that I'm also doing this when one of the sources is from FHLC and the other is from Ancestry, and in this case they should show up on the duplicates list and be merged instead. So let me fix this and update the lists. It will take about an hour for the lists to update. (I'll fix the "computer printouts") problem at the same time.)--Dallan 08:27, 22 August 2009 (EDT)

City Directories [23 August 2009]

Mike provided a great overview of directories under Periodicals (above), but I am running into some dilemmas.

First, the caveat is that my source (repository) for the directories I've referenced is footnote.com or Ancestry.com, and the images don't always include the title page. So, in part I'm dependent on what the repository tells me about the source. Most of the directories for Brooklyn (the location I've been researching) are also available as a microfiche from the Family History library.

Having said that, Mike indicates (wisely, I think) that it doesn't make sense to have a separate directory source for every year. It makes more sense to me to have one book source for all of the editions by a given publisher of the same directory. I'd leave the "year" off the title, and include the range of years for that particular title/publisher in the "Year Range" field.

That would work well if the publisher called the directory the same thing every year. But apparently they didn't. Here's an example: Lain and Company published a Brooklyn directory from 1868 to 1891. For several years before 1868, J. Lain and Co published the directory; after 1892, Lain and Healy published the directory.

• From 1868-1876, the directory is called the "The Brooklyn city and business directory"
• For 1877/78 and 1878/79, the directory is called "The Brooklyn city directory"
• In 1879/1880 (one edition) they revert back to "The Brooklyn City and Business Directory"
• From 1880 to 1882, the title becomes "The Brooklyn Directory"
• From 1883 to 1891, the title becomes "Lain's Brooklyn Directory"
• To make matters even more complicated, when Lain becomes Lain and Healy, the directory continues to be called "Lain's Brooklyn Directory" from 1892-1894.
• And just a side note, prior to the 1860s, there is much more competition in the Brooklyn directory market; several years have multiple directories, each one with a different publisher and slightly different focus (and title).

So here's my comments on the above mess. I understand that the correct *citation* requires that I give the exact name of the directory for the year I am refering to. But fundamentally I don't really care what Lain and Co called the directory; all I really need to know is that it is the Brooklyn directory for a given period. OK, I suppose I care a little, just 'cause then I'll be sure I'm asking for the correct microfiche from the FHL if I want to check it out for myself... (particularly if I needed to research one of the early years where there are multiple directories).

I see several options:

1. I can create five source pages, each one a "book" type, one for each title Lain and Co published. I'd suggest following FHL practice and putting the publisher in the author field, so that there is a difference in the source name when the same title is used by a different publisher.
2. I could create different sources for each year the Brooklyn directory is published, with the actual unique title that references the year. (At least, I'd start with creating a different source for each year I have access to or that I've referenced in a citation...)
3. I could create one source page for ALL the Brooklyn Directories, no matter who published them or what their title is, and just list all the details in the text field; anyone citing the source would have to select/reference the specific edition they used. This kind of fits into Dallan's "subpage" selection discussed elsewhere? It also defines this source purely geographically (all directories for the same location have one source). This looks a bit like the FHL catalog entry for the Brooklyn directories (see link above).
4. I could create one source page for each publisher of the directory (whatever the directory is called), using the most common title or the first title used by that publisher as the title, and then reference other titles and years used in the text. Again, the person who uses this source would be responsible for adding or copying the details that apply to the specific directory/year/edition they cited into the details of their specific source citation (as they must do by adding page number, for example).

Comments? Consensus?--kennebec1 10:09, 22 August 2009 (EDT)

For the sake of ease of use and expendiency, I'd go with #4 and use the most common title. JMHO and your mileage may vary :-) -- Amy (Ajcrow)

I know I refer to library cataloging practices a lot, but they've encountered and dealt with most of these issues long ago. And what Amy says is exactly right. Pick the most common title, call that the "uniform title" (library-speak), and use that as the title of the single Source page. List the variant titles and their years down in the text box, which will assure the user that they're on the right page. You'll still have to do separate pages for the city directories published by Lain & Healy, and those published by West's, and those published by Standard Publishing, but that's unavoidable. --Mike (mksmith) 13:53, 22 August 2009 (EDT)

That's ok, Mike, I learn a lot from your library speak. I've decided I want to be a librarian/archivist/researcher when I grow up. :-) Of course, its a bit late to start that, and it will probably have to wait until I retire... but if I were starting over, that's what I'd want to be when I grew up. -Brenda --kennebec1 14:37, 22 August 2009 (EDT)

I enjoy your library talk, Mike. Helps get me ready for library school staring in a couple of weeks (Kent State). :-) --Amy (Ajcrow) 19:22, 22 August 2009 (EDT)

Libraries (and librarians) aren't what they used to be, you know. My degree, vintage 1970 from North Texas, was "Master of Library Service". We weren't even "scientists" yet. We learned to catalog by being handed a book and actually typing up a set of cards. Now, of course, it's all "information science." Student catalogers today just learn how to logon to OCLC. In the '80s, I could tell things were changing -- and not necessarily for the better -- when our new Director had a for-profit corporate background instead of LS and he immediately began referring to the people who came in the door as "customers" rather than "patrons." I often suspect I retired just in time. . . . --Mike (mksmith) 20:11, 22 August 2009 (EDT)

Renaming rules and Misc types [23 August 2009]

Dallan, You mentioned this and I can't find it - the page has gotten too big. I've been watching as I go and originally scanned all of the 'Other' lists to assess how much of a problem this would be. So far it appears that the program is doing an excellent job of guessing how to handle the Misc types. I am not concerned about this. We're fixing the ones on our assigned lists and choosing a type to avoid the need to guess. But the other million that have not been used or edited should be pretty good. --Judy (jlanoux) 13:16, 22 August 2009 (EDT)

There's a link on the the main page called WeRelate agent (under the Other lists) that lists about 9000 sample source renamings, most of which are for Misc source types if you're interested in looking at more :-).--Dallan 23:28, 22 August 2009 (EDT)

Ctrl-f in your browser allows you to search for text on a web page. I've started using it a lot on THIS page. ;-) jillaine 06:54, 23 August 2009 (EDT)

Deleting Pages [22 August 2009]

I've come across a couple of pages like Source:Van Slyke: The Olive Tree, which are actually ads for a book, not the book itself. How can I delete the page entirely? I've never done that. -- Amy (Ajcrow) 13:30, 22 August 2009 (EDT)

I'll run take care of that one for you. On the More menu is the Delete option - or not depending on your permissions. If no delete option, use the Speedy Delete template. I think it is {{Speedy Delete|reason here}} . This puts it on a list for one of the admins to handle. --Judy (jlanoux) 13:36, 22 August 2009 (EDT)

Thanks, Judy. I don't have "Delete" under More, but I do have "Speedy Delete" under Admin. -- Amy (Ajcrow) 13:41, 22 August 2009 (EDT)

When you use the Speedy template, your delete request will appear on that list. --Judy (jlanoux) 15:21, 22 August 2009 (EDT)

Obsolete Sources - Delete? [23 August 2009]

I didn't quite understand the discussion of census indexes above and I'm encountering more of what I call "obsolete" sources. As I understand it we have source pages serving either of both of two reasons:

• To serve as a finding aid it is a catalog of information available to the researcher
• To serve as a source that will be cited by someone on a Person, Family, etc page

Some nifty thing were once created by the FHL and other agencies such as "computer dumps". While they were once lifesavers, they are now gathering dust as we have reached an era where you have access to the computer databases directly from your computer. I am proposing we delete the entries for these things as I cannot see that they serve either of the above purposes. I would include most paper and microfiche indexes in the category of things we can safely delete. Thoughts? --Judy (jlanoux) 14:30, 22 August 2009 (EDT) (who grew up when census was something you cranked)

I do this all the time. If you don't tell, I won't. ;-) Actually, in seriousness, I usually add the link to the existing source page if I can find it. If I can't and it's something that might be in the IGI, I leave it. (And since I hand cranked censuses little more than 10 years ago, I'll just assume you're 25 or so) --Amelia 23:04, 22 August 2009 (EDT)

I presume we leave it "as is" if someone is actually using it as a source? For example, Source:Parish printout of Hampton-in-Arden, Warwick, England, 1599-1877. -- Amy (Ajcrow) 10:06, 23 August 2009 (EDT)

I think we decided to delete anything that should be deleted. The worst that happens is that the user gets a red link, his citation doesn't disappear. --Judy (jlanoux) 10:33, 23 August 2009 (EDT)

See my comment above -- they can't be wholesale deleted because 1) the information to link them back to the original database is in the FHC catalog entry; and 2) the printouts are in the IGI, meaning they'll be cited and it will be hard to match them up otherwise.--Amelia 11:37, 23 August 2009 (EDT)

The Duplicate List - suggestions [22 August 2009]

I've been studying the list...

• There is a huge number which are books on Ancestry that duplicate another entry. The "Ancestry" page has the correct publication information instead of the genealogical society thing, so I would prefer to save it. (the "Ancestry" name goes away with the rename) But the other page has the FHL repository. I tried to copy the FHL link over and create the repository link. It takes way too long to handle the volume we're dealing with. Is it feasible to try to do this with a script? I.e. tranfer the repository record from one source to the other? This would take care of most of the list.
• Have you tried doing a run to delete all of the sources which point to the old AOL "Hometown" which is shut down? Or more generally is there a way to purge all 'dead website' pages? hmmm...since they all go 'somewhere' that probably isn't possible, but you could identify the AOL ones by URL.
• I am working those which have three or more entries including the Ancestry, deleting the extras and all the other true dupes.

--Judy (jlanoux) 16:10, 22 August 2009 (EDT)

Writing a script to copy the repo information from the other source into the ancestry source and then deleting the other source is a great idea I'll write that on Monday so the dups should be gone on Tuesday.

And I'll wite another script to delete the aol hometown sources on Tuesday so they should be gone on Wednesdy.--Dallan 23:42, 22 August 2009 (EDT)

Removing place from title may cause duplicate title [24 August 2009]

Question regarding Source:United States, Alabama. Death Certificate The source title was Alabama Death Certificates and according to the rules I have removed Alabama from the title so Death Certificates is now the title. Is this correct or will problems arise from this title? --Beth 17:41, 22 August 2009 (EDT)

I think that should be fine, as there shouldn't be another page titled Source:Alabama, United States. Death Certificates, which is what the page title will become. --Amy (Ajcrow)

Ah but here's the rub: while the name WILL become unique, in the interim between now and when it's auto-renamed (robot-renamed?) it's not unique. I ran into this trying to fix double-geographed census source pages. Unless I'm missing something. To provide an example:

Right now, a census page might be: United States, New York, Erie, Buffalo. 1880 U.S. Federal Census Schedule with Buffalo, Erie, New York, United States in the Places covered field. In order to get this page ready for appropriate rename, we need to remove the large-to-small geo entry from the Page Title. Once you do that, you end up with just 1880 U.S. Federal Census Schedule in the Page Title. As you can imagine, that would be a duplicate of other pages. And the system won't let you make such a change.

What to do? jillaine 07:05, 23 August 2009 (EDT)

I don't know Jillaine, but I will need to redo some of the source renaming pages on my list if this is true.--Beth 08:06, 23 August 2009 (EDT)

There isn't a problem in the interim since we're not changing the current Source Page title, just the title field. That's what sets up the automatic renaming when Dallan eventually runs the program. From the Source Renaming Project page: "If the new title is incorrect, please edit the source page and set the source type, title, author, place covered and/or place issued fields so that the new title will be generated correctly according to the rules in the table above. You don't need to (and shouldn't) rename the source. It will be renamed according to the rules in the table above next week."

There will be duplicates when we're done, but what we're doing now won't create a duplicate page before Dallan does the automatic renaming. Someone please correct me if I'm wrong! -- Amy (Ajcrow) 08:51, 23 August 2009 (EDT)

Okay, I must have misunderstood. I thought that by leaving the Page Title alone -- e.g.:

United States, New York, Erie, Buffalo. 1880 U.S. Federal Census Schedule

AND having "Buffalo, Erie, New York, United States" in the Places covered field

resulted in an auto-renamed:

Buffalo, Erie, New York, United States. United States, New York, Erie, Buffalo. 1880 U.S. Federal Census

We were seeing a bunch of those, and I *thought* the answer was to edit the old page title name, pulling off the large-to-small geo text. If that's NOT what we should be doing, then what is it we SHOULD be doing in the above circumstances? Thanks. jillaine 13:48, 23 August 2009 (EDT)

I think you're doing it right, but I'm not understanding the problem. And there shouldn't be any duplicates at any point if you do it right. You should be removing the place from the "title" box (which should just have "1880 U.S. Census Population Schedule") and make sure the place covered is correct. The page will be renamed using Place. Title. The entries will stay on your list because the places are reversed between the current title and the new title, but just ignore them. And MoonKnight has gone through and renamed many, many that have this problem, so you might wait a few days on them, too.--Amelia 14:54, 23 August 2009 (EDT)

The problem in the example that I looked at is the source type is not listed as "Government/Church records". The system only adds the place-covered to the if the source type is "Government/Church records". All you have to do is change the source type on these records and they should be fine. Moonknight should have been doing that at the same time he removed the place from the title, but he must have missed some.--Dallan 12:43, 24 August 2009 (EDT)

Leading articles in the subtitle [22 August 2009]

Is it okay to leave the leading articles in the subtitles? --Beth 17:48, 22 August 2009 (EDT)

I don't see where leaving it in would cause a problem. --Amy (Ajcrow) 18:28, 22 August 2009 (EDT)

Links pages [23 August 2009]

Should Sources that are nothing but pages of links such as Source:New York Vital Records Indexes & Genealogy Records on the Internet be deleted? Yes, they are "finding aids," but they aren't something that someone would use as a source, per se. My apologies if this has already been discussed elsewhere.--Amy (Ajcrow) 18:55, 22 August 2009 (EDT)

Is there a way to call the pages "finding aids" or to add them to the items for the future Research Guide to New York? I definitely agree this isn't a source, but seems unnecessary to redo the the work to identify it as a possible "finding aid" in the future. Maybe its a category that can be added? I can't remember what the outcome of the discussion on finding aids/research pages was. -Brenda --kennebec1 22:51, 22 August 2009 (EDT)

I believe it is a legitimate source, albeit to a web article on "finding aids" to genealogical sources in New York City and New York State. Might be more pertinent as an Article, but would need the author's permission.--BobC 23:14, 22 August 2009 (EDT)

It's your call on whether or not to delete them. There are a lot of these "links pages" on the web. In my opinion, if you find one that's especially good or one that's about a fairly narrow or difficult-to-research topic, it's ok to keep it. But I don't think it's worth keeping the run-of-the-mill links pages.--Dallan 02:24, 23 August 2009 (EDT)

Can someone explain to me under what circumstances such a collection of links page would ever be used as a Source? I don't get it. jillaine 07:10, 23 August 2009 (EDT)

Multiple censuses [22 August 2009]

The FHL lumped together the censuses for Nunney, Somerset, England Source:Census returns for Nunney, 1841-1891. It probably should be split into 6 different Source pages (one for each census year), shouldn't it? The Place page for Nunney lists the current Source page as a resource; that page would also have to be edited to reflect the different sources. I'm tempted to save this one until after the renaming project is done. Thoughts? --Amy (Ajcrow) 22:19, 22 August 2009 (EDT)

I don't know at what level we're dividing UK censuses. If Nunney is the right level, then yes, if you're feeling ambitious, then it should be divided up. But honestly I think your effort is better spent on other things, and I suspect that we are not doing UK censuses by town, so it wouldn't be appropriate anyway. (Similar census books for American towns we have to leave as is for this reason.) At some point we may decide to delete such records as being unnecessary in this world of online access to all of the county censuses, but that's a decision for another day.--Amelia 23:10, 22 August 2009 (EDT)

Funeral home records [23 August 2009]

What subject would one select for funeral home records? --Beth 22:45, 22 August 2009 (EDT)

Gov/church or cemetery (if we still have it). They should be named geographically. (I'm assuming if they had an author you wouldn't be asking.)--Amelia 23:11, 22 August 2009 (EDT)

All of this gets rather confusing. Yes, the "Type" is book and the subject is a different animal. So I should select cemetery for subject. --Beth 23:18, 22 August 2009 (EDT)

Oh, sorry, I misunderstood. Yes, I'd think cemetery would be the closest.--Amelia 23:43, 22 August 2009 (EDT)

Beth, I organized the records of an old, gone-out-of-business funeral home for the Library's archives in Dallas some years ago. They're primarily business records, not really like a cemetery transcription. (Name of deceased, next of kin, how many horses on the hearse, who's paying, etc -- and, eventually, what cemetery.) Businesses are generally classified as "institutions" for these purposes, so I think that's the subject I'd go with. --Mike (mksmith) 00:53, 23 August 2009 (EDT)

I see your point, but I think if I'm looking specifically for burial records, I'd like to have the local funeral home records show up. I'm not going to look under "institution", which suggests the insane asylum (unless of course I'd seen this thread, and now I suppose I might...).--Amelia 00:58, 23 August 2009 (EDT)

Thanks for your input Mike, but down here in Alabama no one would label a funeral home an institution. A school yes and the state insane asylum (as Amelia stated) so I am not sure that institution will help many people from around these parts who are searching for funeral home records. These funeral home records do include relatives, information about the casket, information about the church service and cemetery.--Beth 01:29, 23 August 2009 (EDT)

Subtitles showing up? [24 August 2009]

I was just reviewing my list to check yesterday's changes, and I have several books where the subtitle is showing up in the "new" title. But the subtitle isn't in the title field, and no one has edited it in the last few days. Bug? --Amelia 23:42, 22 August 2009 (EDT)

I started running the page titles through the duplicate analysis process this morning. The list starts with 50,000 duplicates. To reduce the list the process tries one of three things:

• If all of the original page titles in a set of duplicates had an FHL # at the end of the page title, I keep the FHL # in the renamed title in order to differentiate them so they won't be duplicates. The FHLC has thousands of pages (representing the majority of the duplicates) that would have the same title otherwise, because they cover the same place and have the same source title field. It's not an optimal solution, but we avoid dealing with the majority of the duplicates this way.
• If the duplicate item has a subtitle, I add it to the page title in an effort to come up with a different page title for that item. Doing this took care of several hundred duplicates.
• If the items have different publication years, I add them to the page title.

The duplicate list represents the titles that are still the same after all of these things have been tried. If you wanted to give me an example I could verify whether the subtitle was added because it would be a duplicate otherwise. I expect that this is what is going on.--Dallan 02:40, 23 August 2009 (EDT)

Will there eventually be a project to whittle down the FHL duplicates? Many of the ones I've come across could (should) be merged into one page. I certainly don't want to tackle it now, though! --Amy (Ajcrow) 08:58, 23 August 2009 (EDT)

Did you do this to titles that have Ancestry duplicates? (I don't see why it would help, but that's the only explanation I have for Source:Olney, James H. A Genealogy of the Descendants of Thomas Olney. If so, will the subtitle go away once the Ancestry title is removed?--Amelia 13:55, 23 August 2009 (EDT)

Yes I did. To make things simple for me, I added the subtitle to all duplicates that had one, in case the title+subtital combination made the title no longer duplicate. The subtitle will go away once the duplicate is resolved. I'll try to remove the Ancestry duplicates by Wednesday morning.--Dallan 12:47, 24 August 2009 (EDT)

Need someone familiar with Canada to edit this page [24 August 2009]

I don't know why this is on my source renaming list, but would someone familiar with Canada help with this one? Not sure why the ISBN # is listed where it is and not sure about the place issued, etc. Source:Wirth, Mary J. Kammermayer Kinder thanks, --Beth 00:28, 23 August 2009 (EDT)

It's done now. (I didn't find it in LoC so I googled it.) "Kammermayer" appears to be the author's maiden name. I've found the ISBN stuck in almost every field available, including the author field. There's not really a place for it (and I don't know that it's particularly useful on WeRelate anyway), but I moved it to the main text box -- which is where I put anything that doesn't fit elsewhere. --Mike (mksmith) 01:04, 23 August 2009 (EDT)

Thanks so much Mike, there is a place for call number in the repository section, but that is not exactly the same is it? --Beth 01:31, 23 August 2009 (EDT)

No, the "International Standard Book Number" was invented by the Bowker Co. (big-time publishers) as an inventory-control device, similar to a stock number. It worked so well, the rest of the world adopted it and it has become a universal identifier for a particular edition of a particular publication. (And there's a parallel system now for serials.) But the ISBN has nothing whatever to do with the author or subject of the book. It's a code for the publisher + book identifier + check digit. By the way, if you ever publish a book -- even a hand-stapled edition produced on your kitchen table -- you can apply to Bowker for your own ISBN root-code. It's all free, and it gives you something to brag about at family get-togethers. --Mike (mksmith) 14:55, 23 August 2009 (EDT)

And don't forget, if you write "ISBN nnn..." in the text of any wiki page, the wiki system automatically turns it into link that takes you to a page of links from which you can order the book (although the Amazon link is broken right now).--Dallan 12:51, 24 August 2009 (EDT)

Type=Manuscript should grab Author Field, no? [15 September 2009]

I would like the type=Manuscript collection to please include in the page title the contents of the Author field. Example:

Source:George Soule of the Mayflower and his descendants

Thanks! jillaine 07:48, 23 August 2009 (EDT)

One of the reasons I don't like this source project is because I wanted to delete every page that didn't have a link and let people create their own under some of kind of rules that someone intelligent could follow.

That said I cannot find a type=Manuscript collection on my drop down list; so where is it? Trying to fix this page Source:Cemetery records of Oklahoma (650801).--Beth 19:52, 14 September 2009 (EDT)

We removed the type "Manuscript" as a result of an earlier discussion. What once were manuscripts are books (published or unpublished) which grabs author.title, government/church records, which will name the page using place. title., or miscellaneous. Misc. being used for the type of manuscript that is "mixed documents/a collection." For the misc., the page name will grab the author/title unless it is a single-place designated source with a "records"-type name. Which Cemetery records of Oklahoma would, generally, be considered. --Brenda (kennebec1) 20:13, 14 September 2009 (EDT)

See "Type=Book" below for Dallan's succinct summation of our very long discussions over books vs records; I think this one is a "Government/Church record" type. But I also think "Misc." (which it is now) ought to be renamed correctly to Place.Title.--Brenda (kennebec1) 20:20, 14 September 2009 (EDT)

Wow! This is definitely not clear. The classification of source pages is a major headache. How does one expect new users to understand any of this? A cemetery record is not necessarily a government or church record. Funeral homes have cemeteries and private families have cemeteries and towns have cemeteries that are not connected to any government entity. Now on my next one I have another question; so I think the easiest thing for everyone is for me to bow out of this project and let the ones who understand it continue. It will probably save y'all more time in the long run. Then when you finish you can update the help pages so everyone may understand the logic behind the designations. --Beth 20:29, 14 September 2009 (EDT)

Beth, I think of the distinction between "Book" (i.e. Author.Title) and "Government/Church record" (i.e. Place.Title) as being less about whether the source is an "official record" (as many, perhaps most, are not - instead they are transcriptions, or online databases, or vital records books, etc), and more about whether the source is going to be understood by someone searching for it as a "geographically-oriented record" - meaning the source is the type of record (like cemeteries listings, vital records, deed abstracts and the like) where the locality is more defining than the author's name. Remember, the purpose of the TYPE field is (at this point) just to drive the format of the creation of the page name. It doesn't affect the citation of the source (which is still author, book if it is a book); nor does it give the stamp of "official approval or official record" to sources titled in this format. The type just affects how a searcher would look for that title.

So for me, a book *about* the cemetery and its history is a BOOK (author.title). A listing of cemeteries and those interred there is a RECORD (place.title). Because if I want to read about the cemetery, I'm looking for a book. If I want to find someone who's buried in Ohio, I'm looking for a record in a particular place (Ohio, or a county or town in Ohio) This may not be the best example to distinguish, but...

There are always exceptions that prove the rule and I know we could easily reopen this lengthy and confusing debate yet again. But the above is my (internal) interpretation of Dallan's final proposal and resolution to this issue.

having said that, I'm going to spend about seven hours tomorrow in the hospital while my Mom has surgery, so I'll take my laptop with me and tackle the next set of dupes. If you want to share the O-S list, and start again at the bottom, feel free... but I respect your frustration, believe me... I've been plenty happy to not deal with the dupes up until now! --Brenda (kennebec1) 09:10, 15 September 2009 (EDT)

Duplicates [24 August 2009]

Can we get a refreshed duplicate list please? I'd like to see if I made a dent in it yet. My analysis of the list is that - aside from the Ancestry ones - they are about half and half. Some are duplicates and I'm keeping the best and deleting the rest. Others are not duplicates and need to have their title redone to make it unique. European records are especially bad for this. The titles are so long the date part gets wacked off. --Judy (jlanoux) 10:30, 23 August 2009 (EDT)

The list should have been updated this morning. I haven't removed the ancestry-fhlc duplicates yet. I'll do that late Monday.--Dallan 12:45, 23 August 2009 (EDT)

I don't think the duplicate list refreshed. The ones I deleted between 1 and 4 Central were still there when I looked at about 9 Central. --Judy (jlanoux) 13:07, 23 August 2009 (EDT)

Just had a thought. I'm using the link I copied to the Project page (so I could find it). Am I perhaps looking at an old list? --Judy (jlanoux) 13:09, 23 August 2009 (EDT)

The list does appear to be refreshed today - thanks --Judy (jlanoux) 10:28, 24 August 2009 (EDT)

1 and 4AM!?? Wow. Ok, let me clarify. I dump the database about 2am or 3am EST (I don't remember exactly), and the duplicate list works off of that dump, along with the family duplicate list and several other lists. So changes you make in the wee hours of the morning won't show up until a day later.--Dallan 12:55, 24 August 2009 (EDT)

WWII "Old Man's Draft" [23 August 2009]

What should we do with the WWII "Old Man's Draft" cards? Unlike the WWI draft cards, each state has a different National Archives microfilm publication number. (For example, Pennsylvania is M1951, Connecticut is M1962, etc.) Do we leave the different states with their own source page like Source:Pennsylvania draft registration cards (old men), born 1877-1897, or do we create one page to cover all of them (and then, hopefully, someone adds text telling the difference between the different states)? I could make the case either way... -- Amy (Ajcrow) 10:38, 23 August 2009 (EDT)

I think, since we decided to have a single page for World War I draft registrations, we ought to do a similar single page for World War II, and just redirect the various state pages to it. At the moment, for privacy reasons, all we have for WWII is the old men -- but at some point in the future, I assume the feds are going to release all the other registrations, so we might as well take care of it in advance. Then, no matter what form the later information takes, it could be tucked into the citation detail, like WWI. --Mike (mksmith) 15:01, 23 August 2009 (EDT)

New Government/Church records - Website source type [24 August 2009]

Because the original Government/Church records topic has grown very long, and because discussion on this topic is also going on under the "Braintree.." topic, I thought it would be best to start a new topic.

I've thought about the three alternatives that have been proposed for handling derivative/unofficial transcriptions of government/church records posted on websites:

1. Use the "Government/Church records" source type
2. Use a new "Government/Church records - Website" source type
3. Add "(Website)" to the title of these sources

The issue with (1) is that even though we might get everyone watching this page to agree to this option, just the fact that we're having such a long discussion on the topic means that we'd likely need to have this same discussion with newcomers, and I don't want to do that.

Option (3) requires the user to make the same distinction as option (2), and I think it's more work to ask the user to type "(Website)" than to select the "Government/Church records - Website" option.

So I've added "Government/Church records - Website" as an option to the source type list, to use for derivative/unofficial transcriptions of government/church records. I haven't updated the help yet; I'll do that tomorrow.--Dallan 12:43, 23 August 2009 (EDT)

Mmm... if we do "Government/Church records - Website" then shouldn't we also be doing "Government/Church records - book"? OR... combine them both to "Government/Church records - Compilation" or "Government/Church records - Transcription"? I just think "Website" is going to be problematic; use it here, and you'll be getting requests to use it elsewhere. I mean, technically, Ancestry.com's images of census records are visible through a *website*. It just seems really messy to me. jillaine 13:43, 23 August 2009 (EDT)

Ditto. :I've followed this discussion closely and I still have no idea what the use of this type is, so I'm intrigued to read your help. Please tell us: 1) why is a website treated differently than every other derivative or unofficial gov/church record (which gets either gov/church or author, depending on whether it has one) just because it's online? 2) do we use this type when it's a transcription of something that exists already, which would change the rule that such sources are supposed to go on the page for the original records? (In which case, we should stop this project right now and redefine what source pages do, because this is a fundamental change); 3) Do we use it for Ancestry/NEHGS/HQ, etc. and if not, what rule is it you're setting to make that obvious to people? 4) Or do we use it only when we can't figure out what the original records are? I can't imagine how many times we're going to have to explain this to people; it's way more confusing than just telling people that everything that reflects a government record gets the government record type.--Amelia 13:50, 23 August 2009 (EDT)

It seems to me that the primary distinction we're trying to make is between that of a government- or church-produced/published "original" (which includes an image of same, or some sort of electronic digitization of same) vs. a once- or twice-removed human-created compilation (ala "tan books" in Massachusetts published by individuals or societies) or transcript of an original record (i.e., "derivative" and ala some of the USGenWeb or other web sites on which people type up lists of their own). Before I say anything further, are we in agreement about this distinction? Is this the problem we're trying to solve? jillaine 13:59, 23 August 2009 (EDT)

If that's what we're trying to do, then we need to redefine what we're doing. As in go back to the beginning and redefine what a source page is. Because the tan books and census transcriptions and many other similar records are currently treated just like "official" government records -- if something is supposed to be the same, it gets the same source page, and that includes transcriptions, website or not.--Amelia 14:14, 23 August 2009 (EDT)

When did transcriptions become the same as the original source? I've had numerous books that are transcriptions of various records and categorized them as "Book". -- Amy (Ajcrow)

You're doing it right, Amy. Transcriptions aren't the same as original sources. A book is a book. (I thought we had settled that. . . .) --Mike (mksmith) 15:03, 23 August 2009 (EDT)

We're in so much trouble. It's been the rule since we defined what source PAGES (not sources) are that if something is supposed to be the same it gets the same page. So, many transcriptions that only attempt to reproduce another record set are just listed as a link on the source page for the original. Most books do not do this because they pick and choose records or combine them with others, but the distinction we used to have is muddied by the new "if it has an author, use the book format" rule. So you're probably fine. But please don't do use the book format with census transcriptions or similar transcriptions of common records. It just creates extra junk and removes all the wiki benefit.--Amelia 15:10, 23 August 2009 (EDT)

How does it "create junk and remove all the wiki benefit" to create a source page for a source? -- Amy (Ajcrow) 15:17, 23 August 2009 (EDT)

This isn't about creating or not creating a source page for a source. It's about the criteria used to select (or not) "Government/church records" for a given source page. I understand Amelia to be saying: "use gov/church records whether or not the source is original or derivative" which includes whether or not it's a book compilation such records, an original document you witnessed at a church, or a digitized version of the original you saw on Ancestry. And frankly. I concur. The drop-down menu for Type is not distinguishing "Original" for any other content; why should it here? jillaine 15:27, 23 August 2009 (EDT)

Jillaine is right that we're off topic, but this misunderstanding is evidently at the heart of the reason for the "website" thing. To answer Amy's question, source PAGES are not the same thing as SOURCES. Example: 1850 Census records for Smith County, Whatever state. There are two books in the SLC FHC that have transcriptions of this county. One circulates, one has been filmed. Rootsweb has a transcription on file. And of course the films are available many places online and not. The source page should be for Smith, Whatever, United States. 1850 U.S. Census Population Schedule. On that page, there are three county-specific links under the repository and the general information template that tells you where else to find the 1850 census. You can discuss there the differences between the transcriptions if you need to. The vast majority of people will cite the census just by itself -- I find it difficult to imagine anyone is going to cite one of those books; they might cite the Rootsweb page if they are exceedingly diligent. Those diligent people can note that they read only the transcription in the source comment page. On the other hand, if each transcription gets its own page, there will be five different census pages for this one county. They will not be linked to each other unless someone happens to notice they all exist. Someone looking at the main page looking for where they can find that county for free won't see any of the other information. Someone who finds a new transcription has to learn to create a new page. I would be money that no one will cite three of them unless they are confused. So the extra four pages will clog up the database (mulitply four times 13 censuses times most of the counties in the U.S.) for no absolutely no benefit whatsoever. That's all I have time for right now.--Amelia 15:31, 23 August 2009 (EDT)

Jillaine, I believe the "Type" field 'does' distinguish between original and derivative ("secondary") sources.

• Article (derivative)
• Book (derivative)
• Government / Church records (primary)
• Manuscript Collection (could be original, derivative, or both)
• Newspaper (primary)
• Periodical (derivative)
• Website (derivative) (Here "website" is referring to those sites with transcribed, compiled, etc. records. A site with digitized images would be a repository, not a source.)

Amelia, I, too, do not have time for this. I just cannot see how in good conscience we can take a book of transcribed/abstracted records such as Source:Early marriage records of Fairfield County, Ohio, 1803-1865 and say that it is the same as Fairfield, Ohio, United States. Marriage Records. -- Amy (Ajcrow) 17:34, 23 August 2009 (EDT)

And I don't understand why we are now changing everything that we agreed upon in the past. Frankly, I don't see a valid reason. Books were supposed to be entered via the location format if the content was a derivative of a government or church record. After this is over I will probably have to redo everyone of my source citations; if I actually decide to go through all of that. Not sure that I will choose to do so. --Beth 18:28, 23 August 2009 (EDT)

Ditto. I'm glad I'm not imagining things, Beth. Amy, all I can say is that we're talking about different things, and this format isn't working to resolve the difference. You're reading things into the type list that were never intended to be there when it was set up, based on everything I understood at the time. (But no, the sources you mention are not the same. One is a subset of the county marriage records, the other, which I think is fictional, is all of them. Not the same = not the same page.) Dallan, please just make a rule that's consistent. If you all want to go the separate page for every transcription, reprint, extraction, copy route, I will leave this part of the wiki to the rest of you. There are many other projects under the sun, and I'm not going to spend the time on something for which I see no community benefit. --Amelia 18:57, 23 August 2009 (EDT)

Dallan's New Proposal for Type = G/C [24 August 2009]

I really appreciate everyone's patience with this discussion. I know it's been long and somewhat frustrating, but I hope everyone can hang in there until we're able to come up with a solution that truly works. I am not a genealogist, so I really value the input.

It doesn't sound like the "Government / Church - Website" source type was such a good idea after all.

I think my first mistake was my changing the original rule from government/church records compiled into books get the government/church source type to government/church records authored by a human being (not a government agency get the book source type. I did this because I can't see how to automatically determine for a book that looks like it has a human author whether the book would be classified as a government/church record because it's a compilation of records, or a book because it's more of a history.

I'm still not sure how to distinguish between these types of sources, but let's say that we'll have some sources that are of the wrong type initially, but going forward we're going to classify all types of government/church records, whether online, in books, or originals as government/church records.

I'm going to suggest that we have the following types of source pages:

• Government/Church records
• Articles
• Books - but books that are compilations of g/c records are classified as g/c records
• Periodicals
• Newspapers
• Manuscript collections - do we really need this, or can I merge this with Books?

  ::I would like this changed to just "Manuscripts" and retained for unpublished works; I've used it a fair amount. jillaine 07:11, 24 August 2009 (EDT)

• Websites - but websites that are compilations of g/c records are classified as g/c records

  ::I think this "type" is an apple to the other oranges. A web site is simply a location of a source, not a source itself. Could we move this to a separate topic for discussion? jillaine 07:11, 24 August 2009 (EDT)

Furthermore, government/church records come in the following formats:

• Originals
• Digital images
• Microfilm images
• Online transcriptions

Or how how about just "transcriptions" -- again the "online" is the location of the source. jillaine 07:11, 24 August 2009 (EDT)

• Books

As far as I can tell, the questions are:

1. Do we want to create separate source pages for the originals in an archive, microfilmed images of the originals, and the various websites that post digital images of the originals?
2. If several authors have created books covering the same type of records for the same place, do we want to create a separate source page for each book, or list all of the books on one source page? (If we create a separate source page for each book, we could list the other source pages for the same place-covered and subject in a "more-like-this" list on each source page automatically.)
3. If several people have created web pages covering the same type of records for the same place, do we want to create a separate source page for each website, or list all of the online transcriptions on the same source page?

Alternative A If we assume that the answer to (1) is no, the answer to (2) is separate pages, and the answer to (3) is separate pages...

Be careful about those assumptions. I concur that the answers to #1 and #2 are no; but I'm not sure we're close to agreement about #3. I could be wrong. I'm still uncomfortable putting "websites" in the same string of "types" as the others. jillaine 07:11, 24 August 2009 (EDT)

... then we could have the something like three different source types for government/church records:

See my comments above about web sites; I don't think this should be a separate sub-type. jillaine 07:11, 24 August 2009 (EDT)

In each case the record set title ideally should be of a standard form, like "1900 Census Population Schedule", or "Marriages" or "Vital records".

So the title of an online transcription of Such-and-such county marriages posted at rootsweb would look like

Such-and-such, State, United States. Marriages (rootsweb.ancestry.com)

I think we could implement Alternative A reasonably quickly from where we are now by looking at the repository URL for everything that's been labeled g/c records so far, and determining based upon the URL whether it should be of the "- images" type or the "- online transcriptions" type.

Alternative B If we wanted to have just a single source for a record set covering a particular place and type of record, then we could list the various books and online transcriptions as separate repositories on the single source page. We would need to expand the "Repositories" section quite a bit to include author, title, publisher, etc., and we would need to let people who wanted to say exactly where they found the information on the person/family page select a specific repository from the source page to add to their citation on the person/family page.

Thoughts? I'm also interested in better alternatives. I would like to get this right.--Dallan 23:45, 23 August 2009 (EDT)

I'm beginning to lose heart in this continuing discussion. This is something I learned in my first historiography colloquium in graduate school, and it's something we explain in the first session of every "Beginning Genealogy" class we teach. There are two kinds of sources: "Original" and "not original." Everything is one or the other. Everything. There's the original census ledger created by the enumerator in 1850, and there's the image of that very same ledger on microfilm or digitized on Ancestry. Being exactly equivalent to the original, those latter can be treated as "original" (unless you're so paranoid as to think someone has been doctoring the images). But my copy (transcription) of that census, whether it stays in my notes or whether I publish it in a book or whether I publish it on a website -- and no matter how accurate my rendering might be -- is not the same as the original. It can't be. It never will be. A copy is a copy. Not trying to start a fight or ruffle anyone's feathers here, but that distinction is extremely basic in any kind of social sciences research, including genealogy, and it does not seem a difficult point at all to me. And it's the distinction upon which all else depends in the discussions at WeRelate. We simply are not in agreement about the meaning of words like "book," probably because we assume everyone else means the same thing -- but they don't.

Having said that, there seems to be continuing confusion between medium and content. The dropdown at the top of the Source page lists media. Not what the media might contain. But, in fact, that dropdown mixes "media" and "source type." One type of medium listed is "records" (modified here as "government/church," but that really doesn't matter). I take "records" to mean "original, one-of-a-kind, unique documents, created at approximately the time of the event or relationship which they record." (I'm making the assumption that that is what Dallan meant by the term.) That's actually a "source type," not a "medium," but, again, it's a very standard beginners-class definition of "original record."

Including "book" as another medium on the list might, therefore, have been an unfortunate choice, since "book" is such a hazy word. Binding records into a physical volume, like a deed book, doesn't make it a "book" as opposed to "government/church records." I have to take "book" in the WeRelate Source page usage to also mean a source type -- a "secondary source." And in that sense, it doesn't matter whether it's a physical paper book or a website or some researcher's manuscript notes -- because those are all secondary sources, not original records. And that's probably the only meaningful distinction. The practical reason for denoting a source as a "book" or "journal article" or "newspaper" or "website" is simply so the system will know what pattern to follow when we get around to automatically creating citations -- but that practical motivation has nothing to do with the content of the physical volume.

So it looks to me like the dropdown for "media" perhaps ought to be redesignated as "source type." (Again, not content. Specific content has nothing to do with it.) If you choose "original records," then you can further select from among "government," "church," "interview," "institutional," "personal interview," etc. If you choose "secondary," you can further select from among "published" [meaning generally available in multiple copies, like book or article or website], "manuscript" [which, by definition, means one-of-a-kind and of limited accessibility ], and whatever other physical forms you want to include.

Finally, the dropdown at the bottom of the data-entry fields provides a selection of "content types," which is useful for searching the catalog (I guess) but has little or nothing to do with original vs. secondary, so I don't think any more need be said about it. I think that's all I care to say on this issue, frankly. I can't think of anything else that would be useful. --Mike (mksmith) 10:38, 24 August 2009 (EDT)

I hope this discussion is for the future because I can't deal with any more changes within the scope of this project. We've got to get these lists done and the rename in place. Then we'll have some leisure (ha!) to discuss new possible catalog designs and more time to get it right.

The reason for having a type field is because the different media needed different fields to capture the information needed about them. As I change the type, the fields available magically change - some disappear and new ones appear. If you are going to ignore this need and classify based on content alone then just do away with the type field and end all arguments over type. Leave all the fields available all the time. Problem is, this makes it hard to do citations because the format does vary a bit with the media. I am not enamored of computer-generated citations anyway. I would prefer a user field for citation along with the other and allowing for multiple ones.

Judy's vision:

We've talked about consolidating sources - I'm all for that. I would like to see it apply to books also. One page to carry information about all the various editions and reprints with a citation field for each. For Government records and church records and any other type of records, an enhancement of the current "Availability" scheme would be very useful. One could have pointers to other websites, books, etc. where this material has been covered.

But a book is a book. You may call it a duck, but you can't force it to quack. If we collectively won't acknowledge this, there is no point in trying to computer-generate citations. The discussions about shoehorning things into different categories are ignoring the fact that most of them don't follow the same jurisdictional lines as our geographic sources. Tempest in a teapot!

Re websites: I don't see that we need source pages for websites at all. Google and others have very adequately covered the need to find material on the web. Any page we generate faces immediate obsolesence. I see no added value to having pages for them.

So my vision is we have pages for books, one per holding the information for all editions and printings. On that page a field for citations (multiple records) for the user to select. You could computer-generate a default one if you like.

Then we have pages for records (government, church, funeral home, cemetery, whoever). Records pages have places to record the various forms and availability. Availability includes reference to the courthouse, archives, library where the originals reside along with information about other versions: url for a website, source link for a book, etc. Again, with each of these we allow a field for citation which the user selects when he creates his link.

We have very few pages for original government records. This was my impression when I joined and it is confirmed after surveying all of the lists. I think we may need a project whereby we computer-generate a pages for county records (marriages, deeds, probates, whatever) working from a list of counties (US) for each. Then we have seeded the new-fangled Record source type. Users can, as we go, begin the updating and addition of cross-references. It a growing process, but if everyone helps build the records they actually use, we'll get there.

This would go a long way toward making sources easy to use. And we would be taking steps to help educate those who aren't familiar with how to cite. They would just find the book or record they are using and then pick a citation from the list. A "Source Help Desk" manned by volunteers could help beginners. Back to work on the lists, we're running out of time. --Judy (jlanoux) 11:10, 24 August 2009 (EDT)

Dallan's Monday proposal [24 August 2009]

I'm not seeing any concrete alternative proposals above (which I would be happy to see), so I'm going to propose a new alternative based upon what I think I've read.

• All records (government/church, institutional, etc.) get the "Government/Church records" source type, regardless of the format: originals, digital images, or transcriptions.
• All transcriptions of a record set are listed in the repository section on the single source page for that record set.
• Eventually I'll expand the repository section on the source pages to include the option for human-generated citations for each repository. I'll computer-generate citations when a human-generated one is not entered. Along these lines I'll add the capability to store format (originals, digital images, transcription, etc.) and probably other fields for each repository in the list.
• We remove the "Website" source type

This works for me. I originally proposed the "Government/Church records - Website" because I thought people wanted to distinguish between sources that were original records and sources that were just transcriptions of original records. Hopefully expanding the repository section and listing originals and transcriptions as separate repositories fills this need. I have a few questions:

1. How is a book that is a compilation of government/church records classified? Is it listed as a repository in the general source page for government/church records of that type covering that place, or does it get its own source page with a source type of "Book", or both?
2. How is an online transcription that is a compilation of government/church records classified? Is it listed as a repository in the general source page, does it get its own source page (and if so, what is the source type for that page), or both?
3. If a book is turned into an online database, how is the database classified? Is it listed as a repository in the general source page, does it get its own source page (and if so, what is the source type for that page), or both?
4. What do we call websites that are not transcriptions of government/church records? Do we give them a source type of "Miscellaneous" (which is what they used to be given), or do we get rid of them?

Thoughts on this approach?--Dallan 12:28, 24 August 2009 (EDT)

I have no problem suggesting that strict transcripts or photographic images being listed as repositories on the same source page. As long as the intent is that the transcript is identical in content to the underlying source. But if the transcript is at all a subset or superset, then it needs to be a separate source page. One could argue that if page numbering is different, it should be a different source page, but in the interest of compromise... One could argue if there is a compiler, like a website, the author deserves some credit for their work attempting to make data more accessible for others, but in the interest of compromise...

Published VRs contain additional information, and so are not equivalent, much less identical to, town records like you might see on a FHL microfilm. So I have a problem doing anything other than separate source pages in this case. I could cite the published work, referring to data, and somebody else would read the FHL microfilm forever and never find it. Combining these onto one source page is just plain wrong. Further, I don't believe the published VRs qualify as government/church records since they probably contain information that was not originated by the government or church, and represent the choices of the compiler, not the official records.

I think the types need to change. Like mksmith I think content and media types are mixed. Book and periodical seem to indicate the list is media, and if I interpret government/church records as media, I agree with mksmith, it represents the collection of original records, or possibly their strict transcripts. If it is a published book, then book type would seem to take precedence. Websites could be a repository according the strict qualification above (attempt to faithfully copy another source, only), but otherwise, it should be classified as a website.

If the list is content, then where is the choice for history, family tree, etc?

If the list is to control titling, then make the choices "place-focused", "author-focused", and "title-focused". But how are you going to define those? --Jrich 13:11, 24 August 2009 (EDT)

Amen!

Re first bullet: Can we just name them "Records" to prevent further debate? I see this used for any type of recordset, not just G/C (schools, cemeteries, funeral homes, fraternal organizations...)

Re #1: A book is a book. One still needs pages for Books, Periodicals, Newspapers to store the unique sources that these fields require. However, on the Record page for 'Fayette, Kentucky, United States. Deeds' one would be allowed to list the source page for the Jillson book as one of the "availability" listings along with GenWeb Sites, Ancestry, Aunt Susie's website, or any place else someone claims to have found Deeds for this county.

Re #2 We have already agreed that Ancestry, HQ, FamilySearch, etc are Repositories and not 'website' types. One could argue that we could have repository pages for GenWeb and its competitors. Just one Repository page per site and not one page for each subpage within the site, which is happening now. I have also seen some excellent privately done sites that would qualify as repositories in my mind.

Re #3 I would be noted on the source page for the book with its own citation line. It may reside on one of the Repository webpages noted above.

Re #4 If we don't have a website source type, we don't have to call them anything. Get rid of ones that don't fit the structure. We seem to be ready to narrow our vision of source pages from search agent to something closer to 'things one would cite' judging from the comments made over the past few days. I don't see up trying to duplicate FamilySearch's mission of having 'everything a genealogical researcher might ever want to look at' nor a new Google. Those agencies have done a great job, we can leave it to them. I realize this is different from the idea you started with.

The beauty is that we still let people make entries for where these things are available. As it builds, it becomes a finding help, while still maintaining citation integrity. And, we have far fewer Source pages to maintain in the future. People can enter websites without us having to maintain a whole page for this. --Judy (jlanoux) 13:26, 24 August 2009 (EDT)

I"m way to busy to read and digest this today (returning from the vacation that made it possible to be involved last week), but I'm going to chime in and say I wholeheartedly agree with Dallan's Monday proposal. However, I don't think we need to create just one source page for, say Fayette County Marriages, if there are three books of them that cover different time periods/focuses. Those are all sufficiently different to have them be different pages. I think this deals with most of the concern about distinguishing originals from copies -- except for the censuses and cemeteries, most transcriptions of originals I've seen do not aspire to be full exact copies of an entire public agency record set.

On Dallan's questions, 1) Compilation books get their own source pages, named like gov/church if appropriate. 2) Websites that combine several other sources get their own source page; 3) Books turned into databases do not get second source pages; they get a link as a repository and note under usage tips explaining the differences and how to cite; 4) I thought we had already junked most of the websites that were family trees; the ones that remain were judged to be useful. In which case, they can be misc (website has too much baggage and confusion).--Amelia 14:07, 24 August 2009 (EDT)

Monday afternoon proposal [25 August 2009]

I think we're close!

Let me clarify the main purpose for source type: to give the source page a title that is:

• unique, and
• easily findable by entering the first few words of the source page title into the source drop-down list.

We need some way to distinguish between:

• place-title page titles: sources that are geographically oriented
• author-title page titles: sources that have an author
• title-place-issued page titles: used for newspapers so we can create unique page titles for newspapers named just "The Herald".
• title-only page titles: used only when none of the above page title rules can be applied.

As a secondary purpose, the source type is also used to determine which fields to display on the source page, but I've add so many fields lately to the various source types that this isn't that important anymore -- we could just list all possible fields on all source types.

Hands clapping!!! jillaine 20:00, 24 August 2009 (EDT)

As a tertiary purpose, the source type should help the system to automatically-generate a source-citation when possible.

Just to be clear, when I say "repository section of the source page", I mean the list of places on the source page where the source can be found. I don't mean Repository pages. So what I'm assuming here is that Ancestry, HQ, and any transcriptions found on rootsweb or government archives or where-ever would be listed as places where the source can be found in the repository section of the source page. We would not create Repository pages for Rootsweb or where-ever as part of the renaming.

So could have the following source types:

• Records: use place-title format
• Book: use author-title format
• Article: use author-title format
• Periodical: use author-title format
  • periodicals don't usually have authors; this should probably be publisher-title jillaine 16:43, 24 August 2009 (EDT)
• Manuscript: use author-title format
• Newspaper: use title-place-issued format
• Miscellaneous: use title-only format;

Websites containing transcriptions of original record sets should be listed in the repositories section of "Records" sources and would not get their own source pages. Websites containing transcriptions of a variety of different types of records would get their own source pages naturally because we wouldn't already have a source page for that collection of records. The source type in this case would be "Records", the title field would follow the web page title; e.g., "Parish records and tax lists, 1850-1870", and the page title would follow the place-title format.

"Place-title" for a website? Where do we get the "place" from? --Mike (mksmith) 18:21, 24 August 2009 (EDT)

Mike... not the place of the web site's host, the place that the content of the web site is focusing on; e.g., if the web site contains a page on which a transcription of Charlestown, MA BMD records resides. Then the place is Charlestown, etc... jillaine 20:00, 24 August 2009 (EDT)

*sigh* I must be short on sleep. --Mike (mksmith) 09:57, 25 August 2009 (EDT)

Books containing geographically-oriented records would get their own get their own source page.

Non-geographically-oriented records wouldn't have a place-covered field and so the page title would be simply the title of the record set (since place-covered would be empty).

The substantive differences between this proposal and what we said last year are:

• Books: do we create a separate source page for each edition of a book? Last year we said that we would list all editions on the same source page. Now we're saying that each edition can have its own source page.
• Books: Do books that are just compilations of records get titled using author-title or place-title? Last year we said that they would be titled using place-title. Are we now saying that they would be titled using author-title, with a reference to the book from a place-title "record" source?

I would prefer that these (books that are compilations of govt or church records) get the place-title format. jillaine 20:00, 24 August 2009 (EDT)

• Websites: websites that cannot be called: records, book, article, periodical, manuscript, or newspaper source type -- in other words, websites that you would not normally cite, should be removed. The website source type gets removed.

Or use Miscellaneous for those worth holding onto jillaine 20:00, 24 August 2009 (EDT)

I'm going to list all sources that are of type "Website" sometime tomorrow so that we can review them and decide what to do about them.--Dallan 01:37, 25 August 2009 (EDT)

• the repositories section of the source page will eventually have room for a human-generated source citation.

I did not participate in the user interface redesign. Did y'all discuss format for the Source page? If you haven't reached final-final on that, I'd like to suggest (especially after THIS discussion) a redesign of the Source page. jillaine 20:00, 24 August 2009 (EDT)

We haven't discussed that, but I agree it would be a good idea, after we redesign person and family pages.--Dallan 01:37, 25 August 2009 (EDT)

The only remaining question is whether books that are compilations of records be titled using place-title format or be titled using author-title format with a reference to the book from the "Records" source page that is titled using place-title format.

If we say that books that are compilations of records are titled using author-title format, then to keep things simple I would suggest that all online transcriptions of records, whether from a book or from originals (because it's often pretty difficult to tell the difference) be listed in the repositories section of the place-title'd "Records" source page. That will make online databases compiled from books more easily findable. The exact provenance of the online transcription, if known, can be part of the human-generated citation that will be part of each entry in the repository section of the source page.

If we say that books that are compilations of records get titled using the place-title format, then I think we can list online databases compiled from those books in the repositories sections of those source pages.

We already do. For example, a book, whether it's a history or a compilation of VR, might have the following Repositories listed on the same source page:

• FHL
• Ancestry.com
• Google Books
• Open Library
• US GenWeb

My preference is to title books that are record compilations using the place-title format. They would get separate source pages than the source page for the originals. This makes the page titles for all record sets look similar, and makes these books easily findable by entering the place covered in the source drop-down box. It also results in a simple rule: records, whether originals, online, or compiled into books, get the Records source type.

I like this last paragraph alot. I never thought books containing compilations of records should use the Author.Title format because in most cases, the author is unknown, but the place is (i.e., Vital Records of Yarmouth); I don't care who the compiler is; it does not make it easier for me to find. jillaine 16:50, 24 August 2009 (EDT)

Thoughts on this?--Dallan 15:05, 24 August 2009 (EDT)

I still can't see any logic in having books handled differently depending on someone's judgement of what it contains. Many books' coverage does not stay within tidy little county boundaries. You will rely on someone's judgement again to list the right place. It doesn't sound like it is easier to find to me. I don't think that is workable from what I'm seeing going through the lists.

Judy, it won't be up to someone's judgment; it will be based on the title of the book; e.g., Birth, Marriage and Death Records of Charlestown, MA. I think that's pretty clear that the book contains geographically-focused records-- whether original or derivative doesn't matter in this case. jillaine 16:55, 24 August 2009 (EDT)

everyone seems to think of nice tidy little grographic divisions like towns when they talk about this rule. But it has to work for messy ones too. I've seen a lot of source records in the past week. Maybe a book called "The Old Outback" covers parts of three counties. Maybe a book covers an ancient county name which now lies in two or three states. A geographic rule for Books won't work. And there isn't any reason to try to force books to fit that mold. The normal book format works just fine as is.--Judy (jlanoux) 17:13, 24 August 2009 (EDT)

Judy, I understand Dallan's proposal to mean that books will use Author. Title format unless it's really clear that the book is geographically focused. Your example of "The Old Outbook" is not that clear, and so would be Author. Title. But "Birth, Marriage and Death Records of Charlestown, MA" IS. I've also seen a lot of sources this last week, and I haven't run into any ambiguity. Just frustration that I was supposed to use "Author. Title" for these Charlestown records (compiled in a book). jillaine 17:25, 24 August 2009 (EDT)

Been revisiting my Source Rename list, and I think I see where Judy might be concerned. If we're not careful, people might convert History of Sudbury from Author.Title to Place.Title. And I would concur with her, that's probably not a good idea. Judy, would it be better if we more explicit. Book is Author.Title unless the book is a compilation of church or govt records and then it's Place.Title? jillaine 18:21, 24 August 2009 (EDT)

Yes, you're getting closer to my concern. Any time the rule says "except if" you are asking someone to make a judgement call. If the book has ONE record in it, they might decide to use place format. And I've seen a lot of sources this week where someone decided to name every little village in the place covered. So the title goes to whichever one happens to be named first. This rule is going to hide a lot of books.

(Keep in mind that some of those pages that have a zillion places or (as I've found more common) a zillion surnames may have been generated when the FHL records were imported. jillaine 20:00, 24 August 2009 (EDT))

It appears that you and I have different real-life experiences. I find books without an author of any flavor to be rare. And published books of records often do not conform to current geographic place names. Perhaps if the rule said, "if there is no author, compiler, editor, etc. listed"? But you can see how iffy things get. Who is going to be making these decisions? Who is going to be reviewing them for accuracy? I know I sound stubborn, but I've spent most of the past 15 years cleaning up databases. And I've spent the past few months reviewing gedcom files. People put crazy stuff in data fields. They have a lot more imagination than you or I.--Judy (jlanoux) 19:24, 24 August 2009 (EDT)

Ah, but I don't think the creation of new Source pages is going to be anywhere near the problem that say, creation of new person pages from uploaded GEDCOMs is in terms of generating "junk" data. We have MySources to prevent the auto-creation of new Source pages. That's one thing. And I'm not talking about books without an author. In fact the ones that bug me are "books" WITH an author that are compilations of vital records, and the important piece for finding and citing is the place and title, NOT the author. And yes, we need to keep "except when" cases to a minimum. jillaine 20:00, 24 August 2009 (EDT)

So what if we say that a book is listed as author-title unless its a compilation of records for a single place, in which case its listed as place-title? We're asking people to make a judgment call, and so we're likely to have some sources listed twice, once as author-title and once as place-title. It seems (at least to me) that the benefit of making place-oriented sources more findable is worth having some sources listed twice.--Dallan 01:19, 25 August 2009 (EDT)

I envisioned having Record type pages, list all books that cover the area with a link to the book's source page. Why can't someone find the book if this is the case? I was attempting to satify those who wanted books available from a geographic listing.

Judy, you can still have this. I don't see how Dallan's proposal would prevent you from generating this. The "Places covered" field will still be useful for this. jillaine 16:57, 24 August 2009 (EDT)

I agree that you could still have these, and they would still be useful. If we title books that are record compilations using the place-title format, we don't have to rely as much upon someone creating them for people to find the relevant sources though. The source citations for these books would use the standard citation format; place-title would just be for the source page title.--Dallan 17:21, 24 August 2009 (EDT)

--Judy (jlanoux) 16:29, 24 August 2009 (EDT)

I forget to comment on the book-edition part you mentioned. I would like to propose that we have a single page for each book and have a mechanism for sorting out the various editions and reprints and storing them with correct citation for each. I think this would be easier for the user because many people don't understand these issues. He could just pick the one that matches what he has. My hope was that whatever you use on record pages to sort out the various formats (enhanced repository?) would be used for this purpose on book pages. The advantage is that all of the editions and reprints can be sorted out in one place. Is this feasible? --Judy (jlanoux) 16:55, 24 August 2009 (EDT)

How interesting. This also reverts a decision that I made after our original discussions last year. (I know, I should stop making decisions and leave things they way they were.) The decision to add publication year to the page title was made when I was thinking of generating citations directly from the source page, but now that we're going to let each item in the repository section have its own citation information, yes, we can rescind that decision and go back to one source page per book, and handle the multiple-publication information by creating multiple entries in the repositories list. I was actually thinking the same thing, but wasn't sure how this dual use of the repositories list would go over with everyone. So I'm glad that you brought it up.--Dallan 17:21, 24 August 2009 (EDT)

I've been studying the "website" paragraph above. I don't think it needs to be this complex in use. If the website contains transcripts of multiple recordsets (most do) there is no need for a website page. A page either exists or can be created for each Record group. The website can be listed as a Repository on each page. The idea was to eliminate pages for websites altogether. --Judy (jlanoux) 17:13, 24 August 2009 (EDT)

Ideally for websites that contain multiple recordsets, I agree that it would be best to create separate sources for each type of record. And I agree we should encourage people to create multiple source pages in this case. But for people who are lazy, which includes me when we have thousands of websites to review, let's also allow creating a single source with the title of the website until someone gets around to splitting it up.--Dallan 01:19, 25 August 2009 (EDT)

I think I love it (although it gets a little confusing to read all the inline comments). I think this is very simple -- and incidentally the rule I thought we had until the new "Author" thing last week! I don't think we should go the one page per book rule on editions, though. Most such genealogy books I'm familiar with are substantially different from one to another. There are exceptions, like Mourt's Relation that Amy asked me about last week, where the "edition" is just repackaging of the same content, but it seems that typically people do a new edition of a genealogy work only if they are adding all the new research in. (And in practice, if we can't tell that it's different, we can just keep just one page until we know differently.) On the website thing, the big ones I've come across for New England records (Dunham-Wilcox, Ray Brown's place) have repository pages already and generally transcribe record sets that are already in the database (Mass and CT VR), and I've been gradually adding the links in. --Amelia 11:37, 25 August 2009 (EDT)

What Type of Type? [24 August 2009]

I was writing the following while Dallan was posting his Monday afternoon proposal. I'm not sure it's relevant anymore. jillaine 16:52, 24 August 2009 (EDT)

In order to respond to Dallan's Monday Proposal, I need to understand what type of type we're speaking about? As long as we continue to mix medium and content into the Type typology (sorry), I think we will remain in trouble. Here's a table; not saying we have to use all of these-- and perhaps we can combine some-- but the point is, I think we really need to distinguish medium from content (and we may decide not to use medium, but at least let's not mix medium with content-- Marshall McLuhan aside!):

Which of these types of Types do we need?

And please remember: we determine the ORIGINAL vs. DERIVATION at the *citation* level, not at the source level.
jillaine 16:30, 24 August 2009 (EDT)

Finding aid [how would this ever be a Source?] -- I don't imagine it ever would be. Working on the renaming & duplicates lists, I'm seeing this used for genealogy textbooks and all sorts of non-source works, not just actual "finding aids." But since they aren't likely to be used as sources on WR, I wouldn't mind disappearing them. --Mike (mksmith) 18:42, 24 August 2009 (EDT)

Whakapapa [24 August 2009]

I have no idea what to do about the long list of what I suspect are lineage books on the duplicates list under Source:Whakapapa (22357), each distinguished only by the number. Anybody here read Maori? --Mike (mksmith) 12:51, 23 August 2009 (EDT)

Mike, I fear you got ignored on this one. People's attention seems to be elsewhere! ;-) I keep bumping into it, so I thought I'd try to help. It took me a few heart beats, but I think I figured it out. Each of those different numbers refers to a record for a different geographical location; for example, lineages for Whakapapa in:

• Rotorua, Bay of Plenty, New Zealand
• Auckland, New Zealand

So if you can figure out the right "typing" (ahem) such that the title was Place. Title. then it would solve the dupes problem.

At least that's how I see it. jillaine 17:36, 24 August 2009 (EDT)

Renaming picking up multiple authors [23 August 2009]

The system is proposing to rename the page Source:Anderson, Robert Charles. The Great Migration: Immigrants to New England, 1634-1635 as Robert Charles Anderson, George F. Sanborn, Melinde Lutz Sanborn. Great Migration: Immigrants to New England, 1634-1635. Why is it picking up the additional authors? -- Amy (Ajcrow) 15:53, 23 August 2009 (EDT)

This happens when each of the author names are NOT on their own line. I fixed this one. jillaine 16:46, 23 August 2009 (EDT)

I just fixed it also Jillaine; that is too funny. Surname is first followed by comma and given name. It should work now. Source renaming seems to be in a foggy mountain breakdown mode. Let's have a duplicate merge party and see if we can kill off that project. <g> --Beth 18:13, 23 August 2009 (EDT)

Ampersand in book titles [24 August 2009]

This is in the way of a heads-up. I've come across several book titles on Source pages today that use an ampersand instead of "and." Unfortunately, databases frequently screw up special characters, so "Ancestors & relatives of a Svanda family" comes out as "Ancestors relatives of a Svanda family." An ampersand is only a typesetters' convention, not a real "word," so it's okay to change them to "and," and we need to do that in correcting a page. --Mike (mksmith) 18:07, 23 August 2009 (EDT)

That's strange; I thought I was converting "&" and "+" in titles to "and" in the source page titles automatically. I'll check that out. You shouldn't have to convert them by hand.--Dallan 13:01, 24 August 2009 (EDT)

Understanding the purpose of source pages [24 August 2009]

I suggest that everyone take a quick browse of WeRelate:Source review and then return to discuss the prevalent issues. Hopefully it will give you a better understanding of the purpose of source page titles on WeRelate and the goals that were implemented in the process of creating them. WeRelate is always evolving so certainly suggestions for improvement are always welcome. I had a very difficult time understanding the concept of source pages on WeRelate and you can clearly see that on the Source review site. I still have hopes to simplify entering sources. Do we just wish to have a minimum of contributors on WeRelate or do we want to make it more user friendly to encourage more contributions from users? There are previous discussions regarding sources, but this is the most recent before this one. --Beth 20:02, 23 August 2009 (EDT)

I just reviewed this page and its associated Talk page. The pertinent DECISION made back in September 2008 (almost a year ago) about Govt and Church records is this:

Marriage and death records [5 September 2008]

Resolution: title all Source pages with geographically-oriented records (includes government, church, cemetery records, etc.) as "place covered (backwards). website/book/record-set title" For example: Source:United States, West Virginia. Marriages 1853-1970. Source pages for most websites are currently already titled with the name of the website, but do not include the place covered by the records. For convenience, you do not have to rename these Source pages. Once the source review is complete, we'll run an automated process to add the place-covered to the Source page title for sources that have a place in the place-covered field.

The only thing that's changed since that time is that recently, Dallan decided to small-to-large for geographic name order instead of "reverse". I see, however, that what this decision above does NOT address is the Type to use. It's only about naming the source, not how to "type" it.

Have there been changes to this since last September?

-- jillaine 22:09, 23 August 2009 (EDT)

I made a change to books, that books with government/church records that were human-authored should be classified as "Book" instead of "Government/Church records", because I wasn't sure how to automatically determine whether to classify our existing human-authored books as "Book" or "Government/Church records" automatically. I also proposed to add a new "Government/Church records - Website" source type a few days ago, because it seemed like people didn't want to give online transcriptions a source type of "Government/Church records".

But in the latest (Monday's) proposal, I'm rescinding these changes and proposing that we go with the original description above.--Dallan 13:09, 24 August 2009 (EDT)

Type Field [23 August 2009]

One last attempt to clarify... We are looking at records of microfilms made my FHL. The type field is where they coded what they were taking pictures of. Books, Periodicals, Newspapers, etc. and yes, they went to courthouses, churches and archives and took pictures of originals. They didn't say originals because they weren't taking pictures of websites when their project started. But that selection does mean originals. --Judy (jlanoux) 20:10, 23 August 2009 (EDT)

Judy, there is just so much information on this page that I cannot remember what you are referencing. Exactly what are you attempting to clarify? --Beth 20:18, 23 August 2009 (EDT)

FHL microfilm etc [26 August 2009]

By the time we get this all under way the FHL microfilm may be defunct. So do we really wish to spend much of our time creating source pages for obsolete data? At the recent BYU conference, David Rencher's statements were noted by the Ancestry Insider. According to this blog; "Many FamilySearch products are nearing the ends of their lifecycles: - PAF, - compact disc products, - IGI, - pedigree resource file (PRF), - paper publications, (7:50 AM Jul 30th from TweetChat ) - microfilm, - UDE [Universal Data Entry is the predecessor to FamilySearch Indexing]".

Also the FHL catalog is updated every half hour. Link to site: [2]--Beth 22:40, 23 August 2009 (EDT)

It's not the data that is going to be obsolete, but the method of delivery (microfilm, CDs, IGI, etc.). This is where, I think, we are getting ourselves into trouble: confusing data with its delivery mechanism. jillaine 07:22, 24 August 2009 (EDT)

Just as an FYI: Although new microfilm is going away -- they're filming pretty much everything digitally now -- I doubt that the current microfilm will be going away anytime soon. For one thing, there are legal issues around whether FamilySearch has the right to digitize and make publicly available a lot of the films in their inventory. Also, many of the films are used so infrequently that it may not make sense to go to the effort to digitize them, at least not in the near future.--Dallan 13:17, 24 August 2009 (EDT)

THANK GOODNESS!! I have quite a few microfilms to order from Salt Lake. Thank You for the clarification. I was getting worried. Debbie Freeman --DFree 13:25, 24 August 2009 (EDT)

They have a goal to "digitize everything in the vault" in the next 10 years, which is why they make these claims. But reality is that they won't be able to digitize some things for legal reasons, and they won't want to digitize other things for frequency-of-use reasons.--Dallan 13:44, 24 August 2009 (EDT)

On another note, I just realized (after I've actually deleted them as not useful many times) that having the film numbers in the text of the source page is handy for searching when I can't find my source. If I accidently missed a word or the words are too common (Bishop's transcripts of X) when getting something from the IGI or film at the library, the fact that I have the film number means I can search for it and have the right one come up right away. Duh. So keeping them is handy.--Amelia 23:57, 25 August 2009 (EDT)

Actually, the FHL #'s in the titles are not film numbers, they're the ID numbers from the catalog itself. I couldn't use film numbers because a single item may span multiple films. However, the film numbers are listed in the text of the source page, so even if we remove FHL #'s from the titles (which we're planning to do), you can search on the film number by entering the film number in the "keywords" search field.--Dallan 11:21, 26 August 2009 (EDT)

Government/Church records *Type* (why?) [24 August 2009]

ok, I think we must be all getting a little bit tired of this project.... :-}

I'm going to try to pull this discussion back to the concrete again, if possible. I think we are all clear there is a difference between official records and derivative records (and that was a great suggestion for an alternative to term "unofficial records"). We seem to have some disagreement as to exactly where that line is, functionally... i.e. when is a vital records book a derivative and when is it an original?

Brenda, I don't think this last is the problem. I think we can (and should) all agree that an original is an original and a book that transcribes what's on the original is derivative; the key area of disagreement seems to be about whether or not a book (or website) that contains a derivative of govt/church records should be Typed as govt or book or web site. jillaine 07:34, 24 August 2009 (EDT)

But let's get back to the reason this problem was brought up (since I was one of the people who brought it up). What should the TYPE be for those databases, websites, or other non-book sources that 1) are not equivalent to any other source and 2) are not "original" government/church records.

Here are some examples (I'll stick with Braintree)

• Source:Vital records, 1845-1900 (728586) is clearly a Government/church records source, yes? It is an FHL microfilm of records from the Town Clerks office. No pages link here, so it has yet to be renamed or processed, but I think we can agree on this?

• Source:Norfolk Co., MA government vital records. This includes Braintree. The source is an FHL microfiche that does not circulate. I *think* it may be a microfiche of pages from NEHGS? See description: "Microfiche of original published in The New England Historical and Genealogical Register, volumes 3, 9, 11, 12, & 13; v. 36, 37, and 38; v. 1, 4, 5 & 8; v. 5, 11, & 16; v. 13; v. 49; v. 5, 6, & 11; v. 8, 9, 12 & 20; v. 3 & 4; and v. 3, respectively." This is an image of a derivative, technically, and it contains records from several Norfolk towns, not just Braintree. But whether or not those looking at the source think it is "good" or "less than perfect" genealogy quality, what would we call it as a TYPE? (Lucky for us, no one has linked here yet, so I don't think it is on a list....) It is definitely not the same data as the first source.

• It turns out there are two (duplicate) sources for Source:Vital Records and Church Records of Braintree, Massachusetts. See also Source:United States, Massachusetts, Norfolk, Braintree. Vital Records and Church Records of Braintree, Massachusetts. As used (by me, anyway), this is a database on NewEnglandAncestors that was created from a compilation of Braintree vital records by an individual researcher. His handwritten collection of material was donated to NEHGS and is available in their library. The database is a derivative of his work, and his work is a derivative and compilation of official records, church records and "private records." Again, this is not the same data as the first source. It is likely to be similar to the data in the 2nd source, as I think Mr. Sprague was one of the contributors to the NEHGS Register that FHL filmed. But I can't guarantee it is the same data.

This source has been used by me and by contributors to Person:Anne Whitmore (1), Person:John Allis (1), and Person:Thomas Meekins (3). It could be called a manuscript, but the database I used isn't exactly a manuscript; it is derived from a manuscript.

• Source:Records of the First Congregational Church, Braintree, Massachusetts, 1711 to 1762 and 1811 to 1859. This is a microfilm of a compilation by Samuel Austin Bates from the original church records, a typescript manuscript (originally published in a Braintree newspaper). It could be put into werelate as a manuscript or book.

(there is another church record source, and another vital records set that doesn't seem to appear on our database, but I'll spare you more examples)

But let's step back from the question of the quality of the derivations used. All these sources are clearly purporting to be vital records, to be a representation of government/church records. All cover a different record set, albeit sometimes overlapping. In this sense they are all the same "type" of record - (perhaps, one where the Title given to the source PAGE is most useful when it has the {geography. Title.) format?)

Their quality surely varies. Only the first is truly "original."

So here's the dilemma as I see it:

• On the one hand, it is easier for the generic werelate user to find and compare sources that relate to vital records in Braintree if they are all similarly TYPE? The *user* of the source (for a citation) has the opportunity/responsibility to clarify whether the source is primary or secondary, presumably something they will know if they have consulted the record. It is also easier to classify a source you are adding (or updating) as a government/church record TYPE if that is what it is, without having to analyze whether or not it is sufficiently original enough to qualify as the "official/primary" record (and if it doesn't qualify as original, then how to classify it?)

• On the other hand, does classifying all or most these sources as Government/Church records TYPE give the unsophisticated user an illusion that these are all primary records of equivalent quality?

I tend to lean toward the first hand, on the theory that this is the most user friendly, for the werelate user. I'd note that all pre 1845 vital records sources for Braintree are not primary records, and they are realistically the only sources available to most researchers.

But I can see the argument of the other hand. I'd like to come up with a way to directly denote sources quality somehow, if we can get around the fact that we only want one source page per original source, although that page may well contain links to a variety of more derivative databases/indexes etc. that purport to represent the same information.

Dallan has proposed a new option: the "government/church records - website," which is trying to split the difference, creating a distinction between what we've called government/church records so far, and what is, for Braintree Vital and Church records, an even more removed from the original derivative -- i.e. a website/database that indexes a non-official compilation of vital records.

Hopefully this helps clarify why we're having this discussion? I definitely don't want to reopen the discussion of the purpose of source pages. What I'd like to figure out is the best way to label the TYPE field for these sources, which as far as I can tell really only drives the formatting of the Source Page Title (i.e. the geography-Title format). I don't think TYPE field is even a search field? So really, the answer may not affect very much in the larger picture of WeRelate?

Thanks - Brenda --kennebec1 23:52, 23 August 2009 (EDT)

For me to understand this topic, somebody with historical perspective (probably Dallan) is going to have to explain why source pages even ask for a type in the first place.

Type has historically been used for two things: (1) determining which fields to display on the source page, and historically (and again in the very near future) (2) determining how to construct a unique title for the source page that can be easily found by entering the first few words of the source page title into the source drop-down list. I have anticipated that in the future it would also be used to determine how to format a source citation (but see below).--Dallan 13:34, 24 August 2009 (EDT)

Trying to start over on my own, it would seem that different sources need a different bibliographic format, so different descriptive information about the source must be collected, and so type appears to be bibliographic type. Based on this, it would appear to have nothing to do with being original/derivative or some of the other issues being discussed here. It would seem to me that the underlying goal here should be collect enough information to describe the source so that readers of WeRelate can unambiguously locate the source. By normal practice, that means generate a reasonably correct bibliographic entry for it.

As a bibliographical type, then, if an item is a book, it should have a type book regardless of its contents, because that is how the bibliographical entry should be formatted. For the WeRelate reader, this will indicate a different approach to finding it than for other types, for example, looking in worldcat or books.google.com, etc. The assumption is that the government/church records type is for census records and other items that cannot properly be described as a book. Thus, for example, asking that author field be added to Government/Church records, as has been done on this page, indicates that boundaries are being blurred because there is no clear definition of whether type is bibliographic or content or whatever.

Much of the discussion of Government/Church Records here seems to be using it as a content type. Judging the content inevitably leads to considerations of source quality issues, hence further digression into the derivative/original issue. It seems to be that this is all based on a misinterpretation of the type field.

--Jrich 10:32, 24 August 2009 (EDT)

I think the problem that started it all was the (my) assumption that the source page could be used to automatically-generate a proper source citation. If we drop that assumption, and say that each repository listed on a source page would have its own citation field, this puts less burden on the source page itself and I'm hoping it resolves the problem. This is what I've proposed in "Dallan's Monday proposal", and I'm hoping to get comments on this. If people like this proposal, I could add the additional fields to the repositories by the end of this week even.--Dallan 13:34, 24 August 2009 (EDT)

FHL Printouts/Extracted question [25 August 2009]

Hello Everyone,

I would like to rename a Genealogy Society of Utah source that is a Printout (i.e. an index) source for a English Parish Church extracted record to "Location. Index to Parish name etc" I would copy over in the body the FHL title to keep the name connection. Suggestions? Alternate names? etc. Thank You, Debbie Freeman--DFree 13:15, 24 August 2009 (EDT)

Changing the source title too much will make it difficult for people to find when searching for it or entering the FHL title into the source title drop-down list. How about adding some clarifying information to the end of the title within parentheses, or adding the clarifying information in the text of the page?--Dallan 13:37, 24 August 2009 (EDT)

--- OK. I will keep the title and just add the location. Or could I as example "Location. Title from FHL catalog (Index)" work OK? I am not sure how many people understand what the word "Printout" means in this context. I will then add in the body that this is an index, just as I am trying to make it clear what the PRO RG is. Thanks. Debbie Freeman --DFree 13:51, 24 August 2009 (EDT)

If you change the source type to "Government/Church records" and put the location in the "place covered" field, then the source page title will become "place-covered. Title from FHL". You could put (Index) in parentheses at the end of the FHL title.--Dallan 15:12, 24 August 2009 (EDT)

Now I am really confused. The instructions (I thought) say to add by hand the location in the title field to Government & church sources. Example would be "Location. Title". So the sources would be grouped together, and help cut down on source confusion with the same name sources. Debbie Freeman --DFree 15:51, 24 August 2009 (EDT)

Debbie, the numerous "computer printouts" on the duplicates list are merely selective database-dumps -- a partial printout of only part of a record collection, especially of collections like bishop's transcripts from various English parishes. Is that the sort of thing you're referring to? As they stand, they sort-of duplicate the actual record collection from which they were carved out -- and not in any useful way -- so I can't see why we would want to keep them at all. --Mike (mksmith) 19:03, 24 August 2009 (EDT)

Yes Mike that is what I am referring to. They are indexes (seem to be) to the connected English record. That is how the FHL has them listed. I though that is what the FHL website search engine used on their website? Do you know of another source that is an index to these records? I do not but would love to know of one. Would you suggest I create a generic source "Rye, Sussex, England baptisms" and someone help me then add somehow all the FHL baptisms to that page? Or create the PRO Source, and someone help me add the FHL printout source they are based on. The only problem is that EE seems to encourage only citing the source you see. Debbie Freeman --DFree 20:39, 24 August 2009 (EDT)

Since they're indexes (in a limited sense), they're "finding aids," not cite-able sources. I can see that they might be useful at the FHL itself -- maybe -- but I can't think of a use for them at WeRelate. Any source they purport to index, the source itself is also in the catalog. I'm seeing a great many of these things and, frankly, I've marked at least a hundred of them for "speedy delete" in the last few days. --Mike (mksmith) 10:25, 25 August 2009 (EDT)

I don't have an answer about whether or not to keep these sources. I don't think these sources, whether you should keep them or not, are worth spending much time on. So please just do something that seems reasonable that doesn't take too much time.

As for the point of confusion -- whether the place should be part of the title field -- I just re-read Help:Source page titles and I can see how it can be thought of that way. But it should say that the source PAGE title is constructed from the place-covered and source title fields for government/church records sources. You don't need to add the place-covered to the source title field, because the system will add it automatically if you set the source type to "Government/Church records". I'll change that right now.--Dallan 00:45, 25 August 2009 (EDT)

Odd Page Title name construction [25 August 2009]

How did the system translate this:

(old): Freeman, Frederick. History of Cape Cod: Annals of Barnstable County

to be renamed as this:

(new): Freeman, Frederick. History of Cape Cod : Annals of the Thirteen Towns of Barnstable County : Annals of the Thirteen Towns of Barnstable County

I think it must have to do with IF part before (:) results in duplicate THEN use Subtitle.

But why did it use the subtitle TWICE?

Please advise.

Thanks!

-- Jillaine jillaine 18:01, 24 August 2009 (EDT)

That's a bug. Thanks for pointing it out! I'll fix it tomorrow and update the duplicates list with the bug fixed tomorrow afternoon.--Dallan 01:46, 25 August 2009 (EDT)

Rhetorical discussions are not helpful to me [25 August 2009]

I have been working on the duplicate merge project for the past 3 days and this discussion is getting harder to understand. My mind does not relate to rhetorical discussions. So could we select some source pages and show how the last suggestions would affect the pages? Some examples of existing source pages.

• Source:United States, Alabama, Lowndes. Marriage Records
  • A citation from this source: Canterberry-Cochrane, in United States, Alabama, Lowndes. Marriage Records, p. 437, 07 Sep 2005, Secondary quality Redding Canterberry to Drucilla Cochran on 12 Mar 1845 in Lowndes, performed by Justice; surety/perf. name: A. M. Simmons/ Ancestry database; Alabama Marriages, 1809-1920 (Selected Counties).

The title of this source page would stay the same. The repository section of the source page will be expanded someday to allow people to enter source citations for each repository. The computer will attempt to generate source citations for each repository based upon fields in the source and fields in the repository section when a human-entered source citation does not exist. But we've decided not to worry too much about generating source citations right now.

• Source:Alabama Baptist
  • Citation:Obituary for William Coker and Bettie Coker, in Alabama Baptist, p. 3, col. 4, 18 Dec 1884, I3 (I3 refers to the image of the obit on the page)

Ideally this newspaper source would include a place-issued. If the place-issued were "Selma, AL", then the source would be renamed to "Alabama Baptist (Selma, AL).

• Source:Walker County Georgia Cemeteries
  • Citation:Garnie L. Coker, in Walker County Georgia Cemeteries, Vol. I, p. 493, Secondary quality p. 82; Trinity Cemetery on 337 at Center Post, near Center Post School. p. 493; Coker, Garnie L. April 1, 1877 August 24, 1966 together with Coker, T. W. December 6, 1884 August 17, 1968

This source is listed as type "Book" without an author, so it would have the same title after the renaming. Ideally the source type would be changed to "Records" under the "Monday afternoon proposal", and then the source page would be titled: Walker, Georgia, United States. Walker County Georgia Cemeteries after the renaming.

• Source:Cemeteries of Texas, Gloria B. Mayfield
  • Citation:Cemeteries of Texas, Gloria B. Mayfield, 1915 database entry, Loraine Cemetery, Mitchell County, TX, Elizabeth Thomas, 5/18/1856, 2/24/15, Sec. 2

I would change the source type to "Records" under the "Monday afternoon proposal", and then the source page would be titled: "Texas, United States. Cemeteries of Texas, Gloria B. Mayfield".

This is a good example of a malformed page title -- even under the old rules. Gloria Mayfield is a person. She's dead now, but she used to live in Nacogdoches. She compiled a whole series of cemetery transcription books, most of them for individual counties, and now her daughter has taken over most of the project. Since this one is an authored book, the page would be titled "Source:Mayfield, Gloria B. Cemeteries of Texas." --Mike (mksmith) 10:38, 25 August 2009 (EDT)

Ok, I understand that this is an authored book, and that is important to know and retain for the CITATION. But for creating the Source Page TITLE, used in the dropdown find box, isn't it more useful to call this a Record type, and have the drop down show it as one of the many records sources for Texas? Is someone searching for a source for a dead person in Texas going to know Gloria Mayfield's name? ----kennebec1 13:59, 25 August 2009 (EDT)

I agree that since this website is a transcription of a book, its title field should be the title of the book: "Cemeteries of Texas" and its author should be "Mayfield, Gloria B." Since the book is a compilation of records for a particular place, it should be given the "Records" source type, which will result in the Source page title becoming "Source:Texas, United States. Cemeteries of Texas". (But the automatically-generated citation will list Gloria as the author.)

This is not a book. I have never seen the book. If you look at the source page you may clearly discern that this is a website. I understand Mike that you assumed that it was a book because you have knowledge of such a book. While the website may contain transcriptions from the book; there are new updates to the website quite often which I doubt were ever published in this book.--Beth 18:42, 25 August 2009 (EDT)

It sounds like it started out as a book. Regardless, I don't think you could go wrong with a title field of "Cemeteries of Texas" and an author of "Mayfield, Gloria B" and a source type of "Records". You could list on the source page that the website repository contains updates that the book did not.--Dallan 19:07, 25 August 2009 (EDT)

• Source:Poss, Faye Stone. Jackson County, Georgia Deed Abstracts, Books E - G, 1808 - 1822
  • Citation: Jackson County Deed Book E, p. 524, in Poss, Faye Stone. Jackson County, Georgia Deed Abstracts, Books E - G, 1808 - 1822, p. 111-112, Secondary quality "Deed Book E, p. 524 Oct. 2, 188, Jackson Co., J. M. C. Montgomery, sheriff of Jackson Co., to Abner Camp of sd. county, $10 for 4,025 acres, sold by virtue of an execution issuing out of Superior Court of Columbia Co. dated Apr. 7, 1809, suit of Eleazor Cumming vs. John Cobbs. Said land levied on being a tract of land granted to Marbury where the following persons live: George Humphries, Abner Camp, William Henley, John Tredwell, Isaac Knight, John Hill, Presley Knight, William Hill, Absalom Awtrey, Elijah Ragsdale, George Moore, Eleazer Lovejoy, John Antheny, Harmon Holt, Shadrack Humphrey, Mary Edwards, Elizabeth Morris, David Smith, Thomas Grace, Peter Hoyle, Jesse Morris, William Spurgins, Widow Coker, Solomon Coker, Richard Cleaton, Samuel Cleaton, & Calby Carter lives containing 4,025 acres, sold at public sale on Oct. 1, 1811 in the Town of Jefferson to the highest bidder, Abner Camp. Signed J. M. C. Montgomery, Sh'ff. Wit: Rowland Thurmond, Thos. Hyde, J. P. Rec. Oct. 5, 1811. (Note: Two page 524's)."

Under the Monday afternoon proposal I would change this source type to "Records" and the page would be titled: "Jackson, Georgia, United States. Jackson County, Georgia Deed Abstracts, Books E - G, 1808 - 1822". The computer-generated citation would (of course) include Faye Stone Poss as the author when we start trying to generate citations.

And this is also a good example of the conflation problem. These aren't even good-faith transcriptions of the deed records -- they're abstracts. Think of this as a finding aid, because if you cited this as a source in a submitted journal article, the editors would tell you to go back and look at the original deeds and see if they really say what you assume they say. Classifying them as "records" is seriously misleading. The book is secondary work, and it has an author, and I would leave it at that. --Mike (mksmith) 10:45, 25 August 2009 (EDT)

Mike, in my opinion books containing abstracts are useful in research and Faye Stone Poss has contributed much to the genealogical world. Of course I wouldn't cite abstracts in a journal article. And just FYI, last week I visited the Jackson County courthouse and made digital images of every deed for which I cited the abstracts so if I ever get through with all of these projects on WeRelate I may possibly have the time to upload the images. --Beth 18:56, 25 August 2009 (EDT)

Again, I acknowledge the inadequacy of the source, but presumably that's clear (or should be made clear) on the source page itself (and it also clear from the title). If I'm looking for sources (or trying to find this source in the drop down box), I would think I would want this to be categorized under records, because it is specifically about and a collection of public records - yes, abstracts of a public records, akin to an index, but there isn't any other content in the book. I'd want it to be of the type "Records" so that the PAGE NAME is set up with the geographic format, not because I believe it to be the actually deeds. Of course, any CITATION generated should obviously include the author.

I understand your point, Mike. But in the end, whether this source (or the one above) is classed as a Book or a Record won't change its fundamental usefulness or lack thereof. As (I think) I understand it, all the classification does at this point is generate a name for the source page. If we let that be its sole purpose, either sources like these should be classed as books, because that is what they are, and the searcher is likely to have the author's name available to facilitate the search/drop down, or they should be classed as Records, because their purpose (however inadequately met) is to record a set of public records in some form, be it abstracts, indexes, databases, transcripts, images, etc, for a specific geographic location, which makes a page name in that format a more logical search mechanism. --kennebec1 13:59, 25 August 2009 (EDT)

Brenda, I'm not just being an obstreperous old geezer about all this, honestly I'm not. But as you just said: either sources like these should be classed as books, because that is what they are. And to me, that's the beginning and the end of it. Books have been around a long time. The author/title construct has been used for centuries to write them, publish them, sell them, house them in libraries, and refer to them in research. Even non-scholars -- even non-readers -- know what a "book" is and that it has an "author" and a "title." Why is it necessary now to throw out that universally understood system and invent another one? WeRelate's search function for "sources" will search the entire body of information on the Source page, including the geographical subdivisions in the "Places Covered" field -- so why does that have to be in the Page title? And only for certain books, not for all, which only complicates it further? It simply makes no sense to me. More than that, it creates unnecessary work. Work that would not have to be done in the first place if we just used the already-existing author and title of all authored works and then added geographical subject headings where useful. I'm planning to run this whole notion of renaming sources past a couple of academically-inclined genealogists I know next week at FGS, and I'm willing to bet their response will be a bewildered look. --Mike (mksmith) 15:04, 25 August 2009 (EDT)

I want to put this issue to bed. It's been discussed for over a year now. We're going to make an exception for books containing record transcriptions because I believe that the benefit of titling source pages for them according to the place-title format outweighs the disadvantages, and I agree that there are some disadvantages. The "Records" source type will include both primary sources as well as derivatives. The quality of the particular source (or lack thereof) can be described on the source page. We've spent so much time on this topic. Let's move on.--Dallan 15:57, 25 August 2009 (EDT)

• Source:United States, Texas, Comanche. Board Church Cemetery
  • Citation: Alma Nance, in United States, Texas, Comanche. Board Church Cemetery, 07 Nov 2008, Secondary quality

This source would stay the same. (Well, it would be renamed as "Comanche, Texas, United States. Board Church Cemetery".)

• Source:Bryan - Lafayette Funeral Home
  • Citation: Mrs. J. R. Coker, in Bryan - Lafayette Funeral Home, Book 2, Vol. 1, page 69, 09 Nov 1928, Primary quality Mrs. J. R. Coker died around 7:00 A. M. on 09 Nov 1928, age 76, address Harrisburg, Georgia. Funeral at Trinity M. E. Church at 2 P. M. Internment: Trinity Cemetery Minister: Rev. L. B Allman

This book source would be renamed: "Fletcher, Benton L. Bryan - Lafayette Funeral Home"

• Source:United States, Texas. Marriages, 1814-1909

This government/church records source would be titled: "Texas, United States. Texas Marriages, 1814-1909"

This is a database located at Ancestry. They got it by copying the index produced by Hunting for Bears, Inc., so it's second-hand. Nicholas Murray, whose company that was, got it from unsupervised college students he paid to index (not transcribe or even abstract) the records in courthouses all over the country, so now it's third-hand -- and also notorious for being rife with errors. And you copying it from Ancestry makes it fourth-hand -- and the information you've got is incomplete and ultimately untrustworthy because of the distance from the original and the problems noted. And you title it "Texas, United States. Texas Marriages, 1814-1909," which gives a known lousy source the color of an official government source, which is what I continue to object to. --Mike (mksmith) 10:56, 25 August 2009 (EDT)

I just edited the Usage Tips with the information Mike provided. Whatever it's categorized on, this kind of information should be added to Usage Tips when someone knows about it. jillaine 12:24, 25 August 2009 (EDT)

Generally, I think the Usage tips and the source citation categorization (primary, secondary, etc.) is probably where these questions of quality should be dealt with. I think it's great that we educate users (Boy! have I been educated!) in the adequacy and inadequacy of sources. I don't think how the Source Page is titled (as a book, or as a geographic place/Record) is going to do that. I *DO* wish for some way of categorizing the odd (and numerous) Ancestry.com and/or NEHGS databases that lump together numerous other sources, or that are indexes/databases alone with no connection to the actual record images. Not so much in the Source page naming, but perhaps in the quality rating, or in the notations for that particular repository for the source.

I presume that if there is a more primary source for Texas marriages for this dataset, then that would be the source page, and then Ancestry's index (and its inadequacies) would be addressed as a repository for that source page? Is that a correct assumption? Did we decide to try to keep the records pages consolidated when possible? --kennebec1 13:59, 25 August 2009 (EDT)

Well, Texas has 254 counties (yes, it really does), so there are 254 repositories of original marriage records. And then the state has put together a master database for the past 50 years or so, which may or may not be faithful to the originals in the courthouses, . . . but we optimistically treat it as though it is. But any genealogist experienced in Texas research knows all this. And anyone with experience working behind the desk in a genealogy library knows about Nicholas Murray. You live and you keep learning; we all do that. But I can practically guarantee that a novice -- or even a novice in Texas research -- is going to look at "Texas, United States. Texas Marriages, 1814-1909" and the first thing through their mind is going to be OFFICIAL SOURCE. Because that's what that title looks like. They aren't going to read the caveats on the page, they're going to simply link to it and assume they've nailed that piece of data. And that feels to me like misrepresentation. It really does. --Mike (mksmith) 15:16, 25 August 2009 (EDT)

Since this source is a transcription of a book (derivative work), we should have a separate source page for the book and its transcription(s). The author on that source page would be the book's author: Hunting for Bears, just as you have done. We could someday reference this source from the repositories section of a source page for the original records, but it should still have its own source page. People who got their information from Ancestry would (should) reference this source page because its title (Texas Marriages, 1814-1909) is the same as the title they see at Ancestry. People who got their information from original records would (should) reference the source for original records. Most people won't know the difference, but I'd bet that having one source for original records entitled Texas, United States. Marriages and a second source for the Ancestry transcription of the Hunting for Bears book entitled Hunting for Bears. Texas Marriages, 1814-1909 would actually make the problem worse, because hardly anyone would know to cite the Hunting for Bears book. At least if we have one source entitled Texas, United States. Marriages for original records and another source entitled Texas, United States. Texas Marriages, 1814-1909 for the Hunting for Bears book, people will be more likely to cite the Hunting for Bears book because the source title more-closely matches what they see at Ancestry. Also, I don't have a problem adding "(originals)" to the title for the original records source or even "(ancestry)" to the title for the Hunting for Bears book, since Ancestry is probably the most common, though not the only, repository for this book and its transcriptions, when the distinction between them is important, as it may be in this case.--Dallan 17:47, 25 August 2009 (EDT)

• Source:Free Genealogy and Family History Online - The USGenWeb Project (usgenweb.com); this page should be deleted. There are no links.

This source would be deleted.

--Beth 20:19, 24 August 2009 (EDT)

--Dallan 02:13, 25 August 2009 (EDT)

Thank you for posting these examples; my mind too is getting overwhelmed with theory and it helps to try to apply to the concrete. --brenda--kennebec1 14:05, 25 August 2009 (EDT)

United States Federal Census pages question [26 August 2009]

Hello Everyone,

Are we separating the census pages as the following. 1) Keeping the Generic 1850 census page as a Community Source page. 2) List any FHL Book that is about the that census as a book, 3) web page indexes on that census linked to the community source page. Is that correct? Thanks for the clarification. Debbie Freeman --DFree 21:12, 24 August 2009 (EDT)

I agree with (1) and (3). If the FHL item for (2) is for a filming of the original census pages, I would add it to the generic 1850 census page. If it's for a book then I would create a separate source page for it, and assign it a source type of "Book" unless it's a compilation of census records for a particular county, then assign it a source type of "Government/Church records" (which will be renamed to just "Records").--Dallan 02:26, 25 August 2009 (EDT)

Dallan, Good Morning. I more than admit this is beyond my computer skills. These are the two sources that need this clarification. What should I do?

Source:United States, Arkansas, Fulton. 1850 U.S. Census Population Schedule

    Fulton, Arkansas, United States. Fulton County, Arkansas : Seventh Census, Free Population Schedules, 1850

I moved the two books from this source onto their own sources, since books should have their own source pages:

• Source:United States, Arkansas, Fulton. Fulton County, Arkansas : Seventh Census, Free Population Schedules, 1850
• Source:1850 Census of north central Arkansas : Conway, Fulton, Independence, Izard, Jackson, Lawrence, Marion, Randolph, Searcy, Van Buren and White counties

Source:United States, Illinois, Fulton. 1850 U.S. Census Population Schedule

    Fulton, Illinois, United States. 1850 Census of Fulton County, Illinois : Census Information Plus Index

I moved the books from this source onto their own sources:

• Source:United States, Illinois, Fulton. 1850 Census of Fulton County, Illinois : Census Information Plus Index
• Source:United States, Illinois, Fulton. 1850 U. S. Census, Fulton County, Illinois

--Dallan 16:57, 25 August 2009 (EDT)

Debbie Freeman --DFree 12:36, 25 August 2009 (EDT)

I disagree, and since I've done several hundred census pages and set up the page naming and the category hierarchy, I'm going to butt in. Census pages by their nature have a standardized form, and they don't use the book rule because to do so would encourage the insanity of having 2, 3, 7, etc. sources per county, most of which are useless. Plus, following the rules I think were just (re) set out above, transcriptions don't get their own source pages. The correct answer to your question Debbie is: 1) the Fulton, AR record was correctly named to begin with. 2) The "North central Arkansas" book needs to be added under repositories for the relevant counties, although if you really want to it can be renamed as a record with place Arkansas and the title. 3) The Fulton, IL records belong in one record titled Source:United States, Illinois, Fulton. 1850 U.S. Census Population Schedule.

I'll go fix.--Amelia 23:48, 25 August 2009 (EDT)

I'm going to bring this up in a "Separate source pages for books question" subtopic of the final proposal.--Dallan 16:18, 26 August 2009 (EDT)

Genealogy.com famous person pages [25 August 2009]

Are we keeping these sources, or if a better source exists use that better sources, and delete that particular Genealogy.com famous person source? Debbie Freeman --DFree 13:53, 25 August 2009 (EDT)

I don't think we need to keep these sources.--Dallan 17:54, 25 August 2009 (EDT)

Final proposal [25 August 2009]

Ok, here's the final (I hope!) proposal. It's the Monday afternoon proposal with some clarifications. I've added some minor questions that still need to be answered below.

Notes:

• Records include government, church, institutional, etc. records in any format: original, microfilm, digital image, and transcriptions.
  • Most records are expected to cover a particular place. If you have a set of records that are not geographically-oriented, leave the place field blank; in this case the Source page title will be the title of the record set.
  • Do not create a source page for a single record or a small set of records. Create a MySource page instead.
  • If you want to create a source for a website that includes transcriptions of multiple types of records, you are encouraged to create separate source pages for each type of record.
• A Book that is a compilation of records from a single place uses the Records source type; otherwise it uses the Book source type.
  • If the author of a Book is unknown, leave the author field blank; in this case the Source page title will be the title of the book (similarly for articles, manuscripts, and miscellaneous items).
• Use Manuscript for unpublished works.
  • Do not create source pages for manuscripts (or other types of source) that are not available for others to review. Create MySource pages instead.
• Use Newspaper for homepages of current newspapers or transcriptions of historic newspapers.
• Use Periodical for homepages of current periodicals or transcriptions of historic periodicals.

• Do not create Source pages for websites that do are not Records, Books, Articles, Manuscripts, Newspapers, or Periodicals. Create MySource pages instead.
• Do not create Source pages for cemeteries unless the Source page is for an online or book transcription of the cemetery records.

Changes

• I'm going to change the Government/Church records source type to Records.
• I'm going to remove the Website source type.

Questions

I've created various source lists to help answer some of the following questions.

Records bullet 3 question [26 August 2009]

Are we back to creating pages for websites? I thought that is what we were trying to avoid. I would word this as "create a page for this recordset only if one does not exist." Then add the website as a resource under "Availability" or "Repository" or whatever we call it.
I do not like having pages for websites for all the reasons you listed. If we create a "Recordset" page for each website's version of the records, they will all have the same name. How will anyone know which is which? I thought we would have one page for "County, State, United States. Marriages". Then, within that, we will have all the various places these are available labelled Original, transcript, abstract, etc. with a place for citation.

We're going to have one page for originals and their transcriptions, and additional pages for books and their transcriptions. If a website is a transcription of original records, then the title field for the source page it appears on should be a generic description of the original records: e.g., 1850 Census Population Schedule, or Marriages. If a website is a transcription of a book, then the title field for the source page it appears on should be the title of the book. When someone is adding a source page for a website, if an existing source page already has the appropriate title, then the website is added as an additional repository for that source page. If an existing source page with an appropriate title cannot be found, then a new source page should be created. Similarly for online databases.--Dallan 18:41, 25 August 2009 (EDT)

I think I'm starting to understand. The website page now becomes the 'master' record page and the title of the page is now the geog+recordset form? Is that correct? If there are two websites covering the same place/recordset then they somehow get merged?

That's what I'm thinking. Merging multiple websites covering the same place/recordset could be done later or as people notice duplicates. I don't think we have an easy way to detect websites that ought to be merged right now.

And the big question? do we have to do all this before the rename this weekend? --Judy (jlanoux) 19:51, 25 August 2009 (EDT)

I don't think so. I'm going to start a new topic on our schedule below, but I'm starting to think that with all that's been uncovered and discussed this week, that maybe we shouldn't try to rush this and post-pone the renaming for a week or two.--Dallan 11:38, 26 August 2009 (EDT)

Place-covered question [26 August 2009]

How big is a single place???

A country, although it is expected that most places covered will be at the state or county level.--Dallan 18:41, 25 August 2009 (EDT)

This is the question we get into with trying to use geography. If the book covers parts of two counties, do you use geographic form and 'promote' it to state? How do you decide? Can we make it clear that we mean present-day place? (I assume that's what we mean.) So let's look at a book of early land records for Fayette County Kentucky. This county has been subdivided several times since then. Do you go to author format? Promote to state level? It isn't clear to me. We will need a lot of examples to illustrate what is meant. --Judy (jlanoux) 19:45, 25 August 2009 (EDT)

I think it's helpful to keep in mind the reason for using place-title format for "record compilations that cover a single place" is that often the place covered is easier to determine than the first author.

For record compilation books that cover multiple places (which are the minority), we classify them as "Book" and title them using author-title format. If it looks like it covers a lot of the state, you could promote it to the state level if you want. I agree this will be someone's judgment call, but the worst case is we end up with two source pages for the same source.

For record compilation books that cover a single place, even a historical place, we classify them as "Records", list the historical place as the place covered and title them using place-title format. Ideally you would list the child counties below the historical place in the place-covered field so that the source will be returned in search results for the other counties.--Dallan 16:15, 26 August 2009 (EDT)

Newspaper and periodical question [27 August 2009]

Suggest wording to the effect that "newspaper" includes the website, transcriptions, clippings, etc....

I'm ok with adding "historical archives", but not "transcriptions" or "clippings", because I don't want to encourage people to create Source pages for individual newspaper clippings. They should create MySource pages for them instead.--Dallan 18:41, 25 August 2009 (EDT)

Sorry, not clear. I would only want a Source Page for the newspaper and for that same source to be used for the various forms. Not to create a new source. --Judy (jlanoux) 19:40, 25 August 2009 (EDT)

I think we're saying the same thing. The reason that I don't want to add the words "transcriptions" or "clippings" to the wording of the newspaper source type is that I don't want people to create separate source pages for a newspaper clipping that they found online somewhere. I could add wording to the effect: "When you want to cite a newspaper article, create a source page for the newspaper and list the article you are citing in the source citation on your person/family page." How about that?--Dallan 16:24, 26 August 2009 (EDT)

I'm not intending to argue, I apologize. I was suggesting that the wording cover the newspaper as well as the website, etc. Wording got garbled when I ran into an edit conflict - my wiki talents are rudimentary. Whatever wording you come up with is fine. --Judy (jlanoux) 19:16, 26 August 2009 (EDT)

I didn't think you were arguing (I'm sorry if I gave that impression). I've also run into more edit conflicts the past several days than in my entire previous experience. Does the new wording that I suggested cover your concern? How about "Create sources for newspapers listing the website of the newspaper. When citing newspaper sources, list the specific article as part of the citation on the person/family page." I think I like this better. Does it address your concern?--Dallan 13:06, 27 August 2009 (EDT)

Same as newspaper above

Same response as above--Dallan 18:41, 25 August 2009 (EDT)

Websites question

Now, the next two bullets (between periodical and Changes) confuse me. (Because I thought we weren't going to have Source pages for websites.) --Judy (jlanoux) 16:37, 25 August 2009 (EDT)

See my comments under Records bullet 3 above. We do want source pages created for certain websites (when the websites contain record transcriptions, book transcriptions, newspaper homepages, etc.), but not for any website.--Dallan 18:41, 25 August 2009 (EDT)

Manuscript collection question [27 August 2009]

Should we rename the Manuscript collection source type to Manuscript, or should we merge it into Book, and use book for both published and unpublished works? We have 72 sources of type Manuscript collection. Some of them list a publisher.

Manuscript collection type. Without a definition this will promote more pointless argument. A manuscript is literally anything that has been produced by hand (as opposed to mechanically printed.) In genealogy, we deal with a lot of these. To me as a genealogist, a compiled work by an author would be a book, whether it was sent to a commercial publisher or typed and stapled, or typed and not stapled. These are sometimes referred to as 'typescripts'. To me a manuscript collection is a box of loose papers - bills, letters, notes, whatever. So, the answer to the question is a matter of definition and preference. For the sake of utility to the genealogist using the site, he might like to see a "Genealogy of the Jones Family" while he may not want to bother sifting through a box of paper - or vice versa. So I would prefer a "Book" type for the former (the page carrying the info about it being not formally published) and reserve "Manuscript Collections" for the ones that really are collections of 'stuff'. --Judy (jlanoux) 14:45, 25 August 2009 (EDT)

Just wanted to point out that Judy's right about the definition of "manuscript" being an individual matter. We've been sitting here (each with our own laptop) arguing about what that word means. I can give you quite different meanings and usages depending on whether you're conversing with an author, a publisher, a librarian, an archivist, or a genealogist more than 60 years old. --Mike (mksmith) 16:14, 25 August 2009 (EDT)

As noted elsewhere, I'm not sure it's worth while to try to closely define the term; let the user call something a manuscript if that's what he thinks it is. If he's wrong, someone can always change it later. But if it lists a publisher, it's probably not a "manuscript" by any definition. But just as people have been happier with "Family tree/history," I'd like to suggest this one be called "Manuscript/Mss. collection." It gets around the question of whether a single page of correspondence is really different from a file cabinet full of notes. (I don't think it is, for our purposes.) --Mike (mksmith) 16:34, 25 August 2009 (EDT)

If we can't define it, why have it? If we're going to keep it, then let's reserve this source type for a loose collection of 'stuff'. But unless someone argues for this, I'm not sure why we don't call these books for our purposes and reduce the list of possible source types.

The point is that we have to agree on a definition before we can have a meaningful discussion, not that it is not possible to define. Just that no two people would use the same definition. --Judy (jlanoux) 19:36, 25 August 2009 (EDT)

What about the boxes of letters, papers, speeches, invoices, etc that normally constitute a manuscript collection? I am ok with not having pages for these, but they do currently have pages if FHL filmed them.

What does "published" mean? When is something "published"? If I print 30 copies on my printer, punch and put them in report covers to hand to my students - is it published? What if I gave the copies to libraries? What if Kinko's does the work? Or the local job printer? Or GPC's vanity press?

If it is in a library or archives is it "available"? And does that mean it is "published"?

Available means that someone else doing genealogy can access it somehow. So being in a library or archive qualifies. I'm trying to avoid having source pages for family bibles in someone's private possession. These should be MySource pages instead.

I can't answer the published question. (My preference actually is to remove the Manuscript type and change the current manuscript sources into books. I added the Manuscript collection at the advice of someone from the ACPL, but I've honestly never understood what it really ought to be used for.)--Dallan 18:41, 25 August 2009 (EDT)

I agree with changing the ones that act like books to books. I don't care much about the collections. It isn't essential that we have them in the catalog. FHL has them in theirs so it's not like someone couldn't find it. If there are only 72 of them, we could work the list once we know what is desired. Then we could retire the type. If someone wants to enter a source he's used, we always have "Miscellaneous". --Judy (jlanoux) 19:36, 25 August 2009 (EDT)

Manuscripts (or manuscript collections) aren't books. May I make a suggestion? Rather than trying to provide a definition of "manuscript" that will satisfy all its varied meanings, I'll try to define by example. Anyone who's interested, take a short browse through the catalog listings on the left-hand side of the website page for the Dallas Public Library Archives Division. I used to run that, so I know what a wide variety of sources are included. The larger/more interesting collections are themselves links to a list of contents. Maybe 2/3 of the collections in the Division include materials potentially of genealogical interest, and almost all of them include "manuscripts" -- but those are about the only things they have in common. --Mike (mksmith) 19:46, 25 August 2009 (EDT)

I'm going to go back to the outermost level, as 5 indents is enough for me... I just received, in a wonderful act of genealogical kindness, a collection of photos, letters, clippings, death certificates, a genealogy chart, and other misc. "stuff" for one of the lines I'm researching. I anticipate that I'll process what I can use, post what I can for other researchers, and ultimately give much of the material to the Wiscasset/Lincoln County Historical Society, as that is the bulk of the family represented. If they don't want it, perhaps I'll see if the NEHGS would like it. This will definitely NOT be a book. But it will represent a real "source" that could be used by others, and that has genealogical relevance. Possibly in this case for just one lineage, and thus belongs as a MySource(not sure what the critical mass of possibly relevant research is for a source vs. a MySource).

But this is a good example of the type of source that would surely be a manuscript type as discussed above. The Maine Maritime Museum, a repository I haven't yet entered into the database, has numerous examples of this type of document collection (for example: invoices, family records, correspondence, business records, for one of the largest and most prominent shipbuilding families in this area). So my vote is that there should be a manuscript source type. These may be "family history" sources vs. strict "genealogical" sources, but I'd really like to see WeRelate encompass this type of research as well.

Having said that, a very quick glance at the list would indicate that many of these items may be more like a self-published intact work, which is definitely very similar to a book and can easily be defined as a book-like "Type." Often the "publisher" is the Genealogical Society of Utah, which just means FHL microfilmed the content, not that it is really a book, so I don't think having a publisher really defines whether the source type is manuscript-book or a manuscript-collection.

In terms of Source type (i.e. Source page name creation), you can call it what ever you'd like, but I think we still need some kind of type that allows for these types of collections to be included as sources, if they are available to be used by other researchers. However, if the preference is to translate all the currently existing manuscripts to other more appropriate source types, I think the vast majority of them can (and probably should) become either Books or Records. In fact, I just moved the one Manuscript (from my original source review list) to the Government/Church records type. -- brenda --kennebec1 13:05, 26 August 2009 (EDT)

Fascinating example. I would never have thought to call such a collection of individual items a Manuscript or even a Manuscript Collection; I'd use Miscellaneous. Seems like a perfect use for Miscellaneous. In my mind, a manuscript is an unpublished set of writings intended for publication. I went looking for definitions; the last three come close to what Brenda is describing:

1. the form of a literary work submitted for publication
2. handwritten book or document [manu=hand; script=writing]
3. any document that is written by hand, as opposed to being printed or reproduced in some other way.
4. A script before it has been published.
5. the novel as it appears on 8½ x 11 paper, typed or printed from a computer. [others would say this is a TYPEscript]
6. unpublished primary sources. The term manuscript encompasses a broad array of documents and records of numerous formats and types.
7. Individual documents or groups of records having historical value or significance that are not "official records" of university departments or offices.
8. A group within private archives such as literary works or personal papers - different from records created by a corporate body.

-- jillaine 13:39, 26 August 2009 (EDT)

I agree with Brenda's thoughts - very well put. But we have fewer than 100 of these Manuscript Collection records in the database and many of them should move to the Books source type. We have clear evidence that people who don't deal with them regularly don't realize what the 'collection' means. I would propose from a practical standpoint that we do away with the source type, not necessarily the sources. Could we live with collections being 'miscellaneous' source type? and welcome them as sources? --Judy (jlanoux) 13:53, 26 August 2009 (EDT)

After reading Mike's and Brenda's comments I finally understand what Manuscript Collections means -- thank you! Given that they don't seem to be used very often, I agree with Jillaine and Judy that we can call them "Miscellaneous". Would someone mind going through the list and changing the types of the existing manuscripts to Miscellaneous, Book, or another appropriate source type? I'll remove the Manuscript collection type tonight or tomorrow morning.--Dallan 17:18, 26 August 2009 (EDT)

I can do this tonight, tho it might delay my getting back to my primary source review list and the "other" list I signed up for. If someone else isn't busy and wants to do the manuscripts, let me know and I'll be happy to return to other source reviewing. I look forward to being able to return to doing genealogy sometime soon! : -) but it is fun to see sources I otherwise might not ever run across...

Oh, one more caveat - since many of these manuscripts are imported from FHL as manuscripts, if there is some one among us who has access to more info than the Catalog entry provides regarding each document, it might be helpful to have that in determining type. - brenda --kennebec1 18:48, 26 August 2009 (EDT)

Brenda, I can give the list a quick glance (I've spent some time looking at the items on it) and change the ones I think are books. Maybe Dallan can change the rest to miscellaneous when he deletes the type. --Judy (jlanoux) 19:19, 26 August 2009 (EDT)

I've finished renaming all those that look like books. All the rest of the Manuscript type can be change to Miscellaneous. --Judy (jlanoux) 19:54, 26 August 2009 (EDT)

Judy, at least one of the "Manuscript collections" you changed to Book was Source:History and pedigree charts of Mayflower family Lowe family - Stanford family, but when I look at the LDS description of it, there's nothing bookish about it; it is in fact a "Microfilm made of manuscript collection (ca.300 leaves)" also described as "History and pedigree charts of Mayflower family & Lowe family - Stanford family, and papers from..." Seems like this falls under the convert-to-Miscellaneous; not to book. Please explain to us what was behind your decision to convert this to a book. Thanks. jillaine 20:40, 26 August 2009 (EDT)

I am sure I got a few wrong. It's a judgement call. Fix any you disagree with, especially since you are familiar with it. --Judy (jlanoux) 21:58, 26 August 2009 (EDT)

Ok, back to the outer level again - I can't count that many colons in small type! I went through most of the remaining sources (non-books) and added more info and converted the clear-to-me ones to Misc. There are still a few that should probably either disappear or have some additional input; if dallan can refresh the list tomorrow, we can look at what is left. I'm too tired to go back through the ones I still was uncertain about now; I can add more info about these ones tomorrow to help clarify the sources.

I am a bit concerned that we're changing the "manuscripts" to misc. at the same time as we are going through a project to rename the misc to something else??? Aren't we just updating in circles here? In the cases where manuscripts became Misc. I was pretty sure they were best defined as "collections" (see above discussion) and not any other type source, so I chose misc. -- brenda --kennebec1 23:25, 26 August 2009 (EDT)

I'll update the lists this afternoon, as soon as I finish adding comments to this page.

As for changing the type to Misc., I'm not planning to actually change the type on the Misc pages, just rename them. So a Misc page that looks like it contains records will be renamed using the place-title format, as if it had been given a records source type; otherwise it will be renamed using the author-title format, as if it had a misc source type. Probably eventually I'll change the source types of sources that are getting renamed using the place-title format to records so the source type will be consistent with the title, but not right away.--Dallan 13:15, 27 August 2009 (EDT)

I created a user page with manuscripts I have questions on/am seeking input on. Jillaine can comment on it, but can others comment on the user page or does it need to be an article? I didn't want to clutter up the talk page. see User:Kennebec1/Source update questions.

I've also started a page with examples of manuscript (collections) that really are that. Take a look to see some examples of this source type... User:Kennebec1/Manuscripts Collections examples -Brenda --kennebec1 23:06, 27 August 2009 (EDT)

List all fields on the source page question [27 August 2009]

Should we list all fields on the source page, regardless of source type? This has relevance to the first question, since including all fields would mean that manuscripts have the publisher and date+place issued fields included.

I think we should because the type may be changed. But maybe put a note that not all boxes must be filled. Or maybe red * next to required ones the way websites do. --Judy (jlanoux) 16:39, 25 August 2009 (EDT)

I agree. Somehow, whatever field you might omit to display for a certain type of material, . . . it's going to turn out to be important for the next item that comes to hand. (And I'm still hoping for that very short field to record number of volumes. Just two characters. No one ever published a multivolume work of more than 99 volumes, except for law sets, and we don't need those. Okay, three characters.) --Mike (mksmith) 16:53, 25 August 2009 (EDT)

I don't want to put red *'s next to certain fields because I don't want to imply that certain fields are required. Title is really the only required field. Even author is not required for books. How about if I bold the label for "expected" fields, based upon source type?--Dallan 18:41, 25 August 2009 (EDT)

"Expected" might be better terminology. But what I wanted to get across to the user is that they don't have to type in every box, however you think best. --Judy (jlanoux) 19:30, 25 August 2009 (EDT)

I want to see every field for every type. Thanks. jillaine 22:21, 25 August 2009 (EDT)

I'll show every field for every type, and come up with some way to highlight "expected" fields.

If you neglect to pick an "availability" after adding a repository to a source page, the system scolds you and highlights the dropdown box in red. Perhaps that same mechanism could be used to highlight which fields a user should at least try to fill (since you don't want it to appear "mandatory") when creating a source. Which fields are highlighted would change depending on which source type you selected: author, title, publisher, place, date for books -- but only the title and place created fields for newspapers, etc. --Mike (mksmith) 21:54, 26 August 2009 (EDT)

That's a good idea. I'll do that.--Dallan 13:17, 27 August 2009 (EDT)

Periodicals question [25 August 2009]

Should we use the title (publisher) or publisher. title format for Periodicals? We have 84 sources of type Periodical.

Either of the last two will work -- but these definitely need to include the creator/publisher (i.e., society, etc). There are far too many identical or very similar titles, and users will try to find it by typing in the title, most of the time. You want them to know that they've found what they were looking for -- or not. --Mike (mksmith) 13:54, 25 August 2009 (EDT)

If people expect to find the periodical by typing the title, which makes sense, then we ought to go with the title (publisher) format, so that the periodical source shows up in the drop-down list when you start typing its title.--Dallan 15:27, 25 August 2009 (EDT)

(Careful, Dallan, because I'm one of those people who expect to see Books and Articles this way, too (in the drop-down menu): with Title first! ;-) jillaine 22:22, 25 August 2009 (EDT) )

Putting (year) at the end of titles question [26 August 2009]

Should we continue to put (year) at the end of source page titles when we want to distinguish between multiple editions of the same book, or should we create a single source page for all editions of a book and list each edition as a separate repository in the repositories section of the source page?

Copied from above: I would like to propose that we have a single page for each book and have a mechanism for sorting out the various editions and reprints and storing them with correct citation for each. I think this would be easier for the user because many people don't understand these issues. He could just pick the one that matches what he has. My hope was that whatever you use on record pages to sort out the various formats (enhanced repository?) would be used for this purpose on book pages. The advantage is that all of the editions and reprints can be sorted out in one place. Is this feasible? --Judy (jlanoux) 16:55, 24 August 2009 (EDT)

Copied from above: I don't think we should go the one page per book rule on editions, though. Most such genealogy books I'm familiar with are substantially different from one to another. There are exceptions, like Mourt's Relation that Amy asked me about last week, where the "edition" is just repackaging of the same content, but it seems that typically people do a new edition of a genealogy work only if they are adding all the new research in. (And in practice, if we can't tell that it's different, we can just keep just one page until we know differently.) On the website thing, the big ones I've come across for New England records (Dunham-Wilcox, Ray Brown's place) have repository pages already and generally transcribe record sets that are already in the database (Mass and CT VR), and I've been gradually adding the links in. --Amelia 11:37, 25 August 2009 (EDT)

WE know, but the typical user may be unaware of different versions of the book. Especially if he is using a reprint, he may not know which edition it belongs to. Yes, the editions are different - that's why we will allow citations for each. But it is far simpler to list all variations editions, reprints, etc. on a single page so the user is aware that these exist. He may even realize that he should look for the later edition. --Judy (jlanoux) 16:43, 25 August 2009 (EDT)

I vote for Title (year), with one page per edition, concurring with Amelia's point above. jillaine 22:24, 25 August 2009 (EDT)

I just realized, until I'm able to expand the repository section to include the additional information for each edition and we allow people to cite a specific repository in their person/family pages, we're going to need to allow separate source pages for different editions. So let's go with separate source pages for different editions and revisit this once we actually have the functionality that would be required to combine them in place.--Dallan 17:29, 26 August 2009 (EDT)

There are details to be worked out for the pub information. I would agree to get this project done with it as is. Tomorrow is another day. --Judy (jlanoux) 17:44, 26 August 2009 (EDT)

Handling our current Miscellaneous sources question [26 August 2009]

I'm modifying how I title our current set of 900,000 "Miscellaneous" to follow the above rules. In particular, Miscellaneous items that have a records-oriented subject, a human author, and a single place-covered are now going to be renamed using the place-title format (instead of the author-title format that has been used for the past few weeks). There are approximately 130,000 sources with a records-oriented subject and a human author. 85% of them cover a single place so they will be renamed using the place-title format. 1% of them do not have a place-covered and 14% cover multiple places so they will be renamed using the author-title format. Does this make sense? I've created a 10% sample of these places here.

Re the lists Dallan has posted. It appears that 15% of the books (if that's a list of books) containing records won't fit the geographic naming rule. This is what I have worried about. I can't see the lists (404 error) so I'm not really sure what it is a list of. --Judy (jlanoux) 16:59, 24 August 2009 (EDT)

I've fixed the links now. 85% of our current "Misc" sources that have records-oriented subjects and human authors have a single place, so they will use the place-title format. 1% of them don't list a place-covered, so they'll use the author-title format. The remaining 14% cover multiple places. They'll use the author-title format, because our rule says that books that are transcriptions of records from multiple places use the author-title format, unless people think that these sources should use the place-title format with the first place covered.--Dallan 18:58, 25 August 2009 (EDT)

Re. the "Misc. - No places covered" list
I picked some titles randomly, as a sample of your sample. Here's the results:

To Keep:

• Milton Rubincam, Records of the Hathornthwaite family of Lancashire, England, 1611-1734. -- This is a ms. which had nothing entered in its "places covered" field. I've fixed it. So this one was on the list by reason of data-omission.
• Richard Warren of the Mayflower and some of his descendants -- "Places covered" field was empty. I've fixed it.
• Sixth reunion of the Huntington family -- Includes some genealogy but "places covered" field was empty. I've fixed it.
• The Warren family historian -- "Newsletter for the interchange of genealogical data and history of the Warren (and variant spellings) families," but "places covered" field was empty. I've fixed it.

To Delete:

• Statesmans' Yearbook -- general purpose ready-ref book. No genealogical application.
• World of learning -- general purpose ready-ref book. No genealogical application.
• How to search a cemetery -- "how-to" book, no sources.
• Scholars, saints, and sufis: Muslim religious institutions in the Middle East since 1500 -- religious anthropology.
• Yearbook (United Church of Canada) -- annual organizational directory, name & contact info only.

Don't know:

• Brigham Young, his wives and family -- Don't know whether this is family history, biography, or hagiography.
• The revelations of the Prophet Joseph Smith: a historical and biographical commentary of the Doctrine and Covenants -- Biography? Theology?

. . . so it looks like many of these titles can be deleted without hesitation as non-sources, or even non-genealogy. Others are good, legitimate, sources which should not have been on the list, but had empty "place covered" fields. And a few toss-ups I couldn't judge. This list could be culled and corrected by hand but I don't see a systematic way to do it. --Mike (mksmith) 19:10, 25 August 2009 (EDT)

Re Miscellaneous - one place covered list: I quickly looked at this list, and I think there are many that should be in the (author. title) format. They are place-named sources, but many are "Histories of" places, not records. If possible, you may want to try to script a finer sort of the sources with a place name, separating those with the word "cemetery(ies)" "burial(s)" "birth(s)" "death(s)" "marriage(s)" "inscription(s)" "will(s)" "deed(s)" "census" "baptism(s)" "records" in the title, and also separating ones with "History of" or "City Directory" in the title. The former are more likely to be Records (place. title.); the latter likely to be books (author. title). Then the question is how many are left, and will they have to be reviewed by hand?

I have, by the way, no understanding at all of the software side of this, so if this is impractical, I apologize. But there definitely needs to be a finer sort before a decision is made on these sources, whether by hand or machine... The other option is to acknowledge they will have to be corrected later as they are used. - Brenda --kennebec1 13:43, 26 August 2009 (EDT)

Keep in mind that some "non-genealogical" sources might still be relevant here. For example, in my article about Spiritual Wife-ism in Colonial Massachusetts, I cite some historical sources that were used to help support the article. I suppose one could argue that these could/should remain MySources, but just wanted to point out that not all non-genealogical sources are irrelevant here. (Besides if we were to really apply such a rule, we'd have to get rid of Wikipedia!!!!!) %-O jillaine 13:44, 26 August 2009 (EDT)

Ok, so it sounds like sources without a place-covered should be listed as author-title if they are to be kept.

For sources with a single place-covered, I'll run some keywords over the titles to filter it. That's a great idea.

I just reviewed a number of multi-place sources (and fixed the link). Some aren't record compilations, and I'll filter them out just as for single place-covered sources, so they'll be listed in author-title format. Many cover two places unnecessarily; for example, they list both the county (correct) and also the state (incorrect, since the source covers the county only). This looks like a bug from when the source pages were created. So if a source lists two places but one is contained within the other, I'll treat it as a single place-covered source. A few list different places in the places-covered field, but only one of the places covered is listed in the title. I think it makes sense to treat these as single place-covered sources as well. A few others cover multiple places where both are listed in the title. These will be listed in author-title format.

So the rule for record-compilation books is classify a compilation of records as Records if it lists a single place-covered in the title; otherwise classify it as Book. It's not perfect, but from what I can tell it will handle the majority of our record-compilation books reasonably. Over 90% of the record-compilation books that we currently have source pages for list a single place in the title.--Dallan 20:27, 26 August 2009 (EDT)

Website sources question [28 August 2009]

Should we delete all of our current "Website" sources, or just those without records-oriented subjects? Maybe we should re-classify website sources with records-oriented subjects as "Records", and delete the rest? We currently have approximately 2200 website sources with records-oriented subjects, 6100 with "Finding aid" as a subject, and 1500 with other subjects. I've created lists of the website sources here.

"Finding aids" are very iffy. Some of the sources on this list are for family and clan societies, some of which include some (or a lot of) records relating to the family, . . . along with family photos, details of DNA projects, and all the rest of the stuff you would also expect to find in any society that holds picnics in a park somewhere. All those latter materials are probably useful, but they're not "sources." But I also see quite a few titles on this list (but still a small percentage) that have been mis-titled, and that are something other than a "finding aid." I suspect people have been construing the term to mean "miscellaneous." (Or "I don't know what to call it.") I think a fast check through the list could identify those that ought to be kept, and re-labeled. And I would vote to discard (as being non-sources) all the rest.

On the question of simply deleting all website sources, or even a large fraction -- I'm not sure we can get away with that. The web is a legitimate medium, like books or periodicals. For many newer family researchers, the web is the source, like it or not. If you want to attract those new to the hobby, you can't simply leave out what they regard as their most important resource. At best, I think we'll find many of those website sources being recreated by new users. --Mike (mksmith) 14:21, 25 August 2009 (EDT)

Do we have to deal with the website issue now? We've already reviewed most everything needed for the rename. Can't we tackle websites after the rename? If you could automatically generate the 'generic' source for common county records we could begin moving websites as repositories and then deleting the source for the site. Those which don't contain records would be deleted. I hate to see the issue of "website as a record type" opened up again. Once someone changes "website" to "Record" we'll never be able to find them to create the work list we need.

If the answer is that someone is willing to review these pages by hand to determine which ones should be kept and which should be deleted -- that's great. I'll keep them listed as Website then. I suspect that only the websites with record-oriented subjects are worth this human attention. The others we may just want to delete.--Dallan 18:05, 25 August 2009 (EDT)

BTW, we're not talking about deleting all website sources. Many of them have already been classified as government/church records when the website-source review project took place last year. That project required manually reviewing 120,000 website sources that our Web crawler had found in 2005 and classifing them. Most of them were deleted because they were for people's individual family trees. Currently we're talking about deleting website sources that people would not be likely to cite, which is a stricter criteria than what we used last year (which was to delete sources that were not useful to most people).--Dallan 18:58, 25 August 2009 (EDT)

While I haven't read every word here in what is turning into a massive subject content, it sounds as if the general consensus about citing websites as valid sources may be eliminated, limited or curtailed. I guess I side with Mike's viewpoints regarding this issue. I agree that most websites may be considered secondary or tertiary in nature when considered as a source for genealogical data; and unproven, questionable or unreliable in the reliability factor. But I feel the web still should be considered as a legitimate repository of a multitude of genealogical source work.

A PEW study from about five years ago showed that the use of the Internet for hobbies has grown exponentially, and that by the date of the study that 24% of general respondents went online to research family history or genealogy. On another research study conducted more recently of genealogists and people researching their family history, "All participants made use of the Internet and participated in some form of genealogical organization either locally or abroad." Genealogical use of the Internet among that group showed that it included a variety of activities such as research in online resources or databases, seeking information in chat rooms or listservs, finding contact information for libraries of archives, and planning research trips.

The organizers of this WeRelate website count on genealogists and family researchers to use this site for some of their research. For WeRelate to ignore or banish most websites from its approved primary source library or to force users to religate the use of website sources solely to the "MySource" library are doing a disservice to researchers.

A few months ago I read a National Genealogical Society article that identified six genealogical standards and research guidelines as good rules to follow for all genealogists and family historians and created an extract of those standards for use of technology in genealogical research. A couple of the more pertinent guidelines include:

• Treat compiled information from on-line sources or digital databases in the same way as other published sources--useful primarily as a guide to locating original records, but not as evidence for a conclusion or assertion.
• Cite sources for data obtained on-line or from digital media with the same care that is appropriate for sources on paper and other traditional media, and enter data into a digital database only when its source can remain associated with it.

If I've overreacted, please excuse. As I stated earlier, I've perused this subject rather than reading every word. --BobC 13:20, 26 August 2009 (EDT)

Bob, you have some good points. I don't believe we want to ban websites. But it is problematic having a source type labelled website. Website are rather amorphous and have lots of pages that contain different types of information. We find that users are creating source pages for single pages within a larger site. This tends to hide the information rather than make it useful. So we propose to list the website on the page with the recordset. That way, you could go to the records page and find perhaps several websites that contain these records or part of them. The same website (in the strict def. - a collection of pages) could appear on many different source pages. We would have the facility for storing a citation for the website as a source for this particular recordset. It gets full credentials, it just wouldn't have separate pages dedicated to the site itself. Does that sound ok? --Judy (jlanoux) 20:06, 26 August 2009 (EDT)

I'm sorry to give the impression that we're banning websites. We're not, and in fact have gone to fairly great lengths to scour the Web looking for good ones. As Judy says, we're listing websites on the pages with the record sets. For websites that are things that one would probably not cite as a source, such as finding aids, a good point has been made that there are other ways to find these websites, and we don't need to include them in the source database.--Dallan 20:45, 26 August 2009 (EDT)

I have been waiting, hoping I had been fooled by the long conversation where it is impossible to tell what is authoritative. But it seems every time website is mentioned, like in the above paragraph, it is limited to being records,or transcripts of other published sources. I believe any such limitation would be a serious mistake. I can give examples of several good websites that are not record-oriented, are not some other type of transcript, and yet I believe are valid citations for the following reasons:

1. they provide data I have not found anywhere else, yet.
2. they cite their sources
3. they cited sources that are out of print and which I understand to be the definitive work on certain families, which I believe by reputation to be reliable but cannot obtain or find in a library within a day's drive of me
4. I cannot with intellectual honesty cite the underlying source, and must cite the website even though the main purpose of the website is a family tree, because I have not seen the underlying source.
5. Many of the authors of these better websites are better genealogists and more reliable than those of the books and manuscripts that I would have no problem citing.

I don't understand why the anti-website decision. The Internet is the future of genealogy. 90% of those self-published manuscripts from earlier in the century that I could cite only exist because at the time they were compiled, there was no Internet. Now those same sources are Internet websites. Yet the old ones are OK, but the new ones are not? --Jrich 21:13, 26 August 2009 (EDT)

Jrich, Dallan just above wrote "I'm sorry to give the impression that we're banning websites. We're not..." There is no "anti-website decision". We're talking about appropriate use of the TYPE field, and where best to include a link to the websites in question. So instead of calling a source that happens to reside on a website Type=website, the Type will be used to describe what's on the website, not that it is a website. jillaine 21:25, 26 August 2009 (EDT)

I didn't see a type for Family tree in the list. Did I miss something? So how to cite a website that is a family tree? In Dallan's paragraph above he said, quoting Judy, "websites on the pages with the record sets", implying if the website isn't a collection of records, it's not wanted. Elsewhere there have been comments about deleting websites that are family tree, only keeping those that are records or transcripts. That's why I am asking for clarification. Are we saying either 1) WeRelate doesn't want people citing Pane-Joyce, Carew-Lunstedt, or other good websites as sources or everybody creates their own MySource for these and instead of one source page, we have multiple MySource pages, one for each person that thinks one of these is worth citing? --Jrich 22:41, 26 August 2009 (EDT)

I understand that this site is still beta, but the speed at which these fundamental policy changes are evolving is surprising. We went 10 days ago from identifying the majority of website sources as "Miscellaneous" types of records to identifying them as "Website" types under the new source page title rules. Okay. Small adjustment. No problem.

Somewhere in the process since then "Website" became a dirty word. Now it seems we're trying to re-categorize these website sources as articles, books, manuscripts, g/c records, newspaper, record sets (whatever that is), and converting others that don't fit neatly into one of these predefined type pegs to "MySources" and deleting the rest.

And how "Finding Aids" became a red-headed step-child with no apparent genealogical source value and worthy of only a pull-the-plug Speedy Delete is untenable. Come on... How about re-categorizing them as "Help Pages" referencing external resources.

Please, Judy, Beth & Jillaine (and other Speedy Deleters) slow down, take a breath, and let's develop a coherent understandable policy we can adjust to, comment on, test, and disseminate adequately. --BobC 01:39, 27 August 2009.

Just for the record, Judy's position was to defer handling of website type until after the renaming project when we could study the list and assess how they should be handled. (I thought Dallan had agreed to do this.) I do not delete websites unless they are bad links or are otherwise inappropriate.--Judy (jlanoux) 18:43, 27 August 2009 (EDT)

Dallan has agreed to this.--Dallan 17:23, 28 August 2009 (EDT)

Bob, I'm actually working (separately) on Research Guides, and I could see that some of these "finding aids" might have an appropriate home there. And I'm happy to review these finding aids to see what's appropriate for including. But if finding aids are sources for genealogical information, I'd like to see examples. Thanks. jillaine 13:20, 27 August 2009 (EDT)

Some examples I can think of off hand might be...

• Guild of One-Name Studies and associated pages on the external website
• Department of Veterans Affairs Gravesite Locator
• Bruce International Surname Study
• Swancott Surname Study
• Royalty Resources
• Matz Surname and associated link
• Knibb(s) One Name Study and associated link
• Free Surname Search: All of the UK
• Fallen Heroes Memorial
• My own series of OIF-OEF Casualty pages for 2001, 2002, 2003 and 2004 (which I've already edited to be changed to "MySource" types for preservation sake
• Many of the pages listed at Web Finding Aids]

Can probably name name others, but that should do as examples. I'm flexible, and as I see the train already rolling down the track I'm preapred to keep my opinion to myself about this subject hereafter and ride along with the rest of you. --BobC 14:44, 27 August 2009

I will help with the human review of website-coded sources. jillaine 22:28, 25 August 2009 (EDT)

I can't do it as a rush job with a deadline, but if you're willing to retain them for a while, I'll also help with websites and with manuscripts. Once we empty them, you can kill the type. Do you have a way to prevent new ones begin added? --Judy (jlanoux) 10:40, 26 August 2009 (EDT)

That would be great - thank you! I'll remove "Website" and "Manuscript collection" as possible source types, which won't affect the existing sources, but won't allow new ones to be created.--Dallan 20:45, 26 August 2009 (EDT)

MYTH: We're banning websites; REALITY: we're trying to TYPE websites [29 August 2009]

What’s been said by Dallan about websites and sources:

• We do want source pages created for certain websites (when the websites contain record transcriptions, book transcriptions, newspaper homepages, etc.), but not for any website. (25 Aug)
• BTW, we're not talking about deleting all website sources... we're talking about deleting website sources that people would not be likely to cite. (25 Aug)
• I'm sorry to give the impression that we're banning websites. We're not... (26 Aug)

Let me try another approach:

All we’re trying to say here is that as a Source TYPE, “website” is too vague, used for too many things, too easily confusing.

What TYPE of SOURCE is on the website? That’s what we want those sources currently Typed as “website” to be changed to.

Is the type of source on the website:

• an article? Use type=article
• derived/transcribed from a book? Use type=book
• transcribed from government, church or cemetery records? Use type=Government/church records
• transcribed from a newspaper? Use type=newspaper
• an online version of a periodical? Use type=periodical
• none of the above? then use type=Miscellaneous

I do see you saying that we may need a type to reflect that the greatest number of likely-to-be-referenced web pages are probably family trees. Family tree/history is at the next level down, on the SUBJECT drop-down menu. So what might the TYPE be? How about:

• Personal research

or

• Family records (the opposite end, if you will, of "government/church records"

Does this reframing help at all?

-- jillaine 13:13, 27 August 2009 (EDT)

Thank you. How about calling a person's legitimate independently researched and sourced website family tree (versus a questionable World Family Tree type collection) or a personally created genealogy listing as an "Article"? It may have an author and copyright, and a "Source" listing would be better able to reflect those pieces of information than a "MySource" page. I basically see those well-written and personally researched family trees as an on-line articles, and they could be identified as "Family Tree/History" in the subject. --BobC 00:35, 28 August 2009 (EDT)

The over-arching rule for websites, or any source for that matter, is:

Keep websites that people are likely to cite.

Another consideration is:

Keep websites that people would want to be referred to in a GenSmarts-like source referral feature.

Someday I want to give people, especially newbies, the ability to ask: where should I go next, either for a specific person or for everyone in their tree, and to get a list of sources they should check. I really like the GenSmarts program, and I think that a community-driven database that includes both offline and online sources is the right way to implement that functionality.

For the websites that you choose to keep, re-classify them as something other than "Website", since this source type is causing confusion and is going away. If a website that you want to keep doesn't fall into one of the other source types, put it in Miscellaneous. Not having done nearly as much genealogy as everyone else watching this page, I'd like to let you all decide which websites fit those rules.

Also, I'm not against people citing websites. The issue is that several years ago we were very proactive and added over 100,000 genealogy-oriented websites to our source catalog. Many of these turned out to be for individual family trees, which in most cases (not all) should be listed as MySources since they're relevant to relatively few people. So last year we removed them. And now we need to remove other websites that don't fit the above rules to avoid cluttering up the source wiki. We're doing similar things with some of the 900,000 sources we imported from the FHLC: if it doesn't look like anyone's going to cite it, and we don't think it would be helpful to list in a GenSmarts-like source referral system, then delete the source.

But going forward if someone wants to cite a website, even one we deleted, that meets the general criteria for Source pages: being relevant to more than a few people (make it a MySource if it's not), they are welcome to create a Source page for it.

The reason that I don't like websites being listed as Miscellaneous and want to list them as g/c records, books, or newspapers whenever possible is that Miscellaneous sources use the author-title format for the source page title. Websites rarely have authors listed, so they'll generally be listed just as "title" only. But website titles aren't like book titles -- they change over time. As I've been going through the website sources that we currently have, often the title that we obtained from the website when we first crawled it several years ago is different than the title it has now. So someone trying to cite the website is going to enter the title it has now, not find an exact (and sometimes even a close) match, and so they're likely to create a new source for the website. So then we end up with two source pages for the website, and over time we'll have even more. But if we can re-classify a website as g/c records, then the place-covered will be added to the source page title. The places covered by these websites don't change as much, so if we classify these websites as g/c records, then the source page titles will use the place-title format and they'll be easier for others to find later.

Finally, for people doing the deleting please don't delete sources that people are actually linking to.

Does this sound ok?--Dallan 14:02, 27 August 2009 (EDT)

Regarding Jilliane's comment re family records. In my experience, good family web pages are much more like a collection (a manuscript collection) than anything else. They have everything from photos to letters to family trees to records of military service.

This type of site (source) can become miscellaneous type for now, as we have for now called other "collections." But fundamentally, source TYPE drives 2 things: 1) how the source page will be titled, and 2) what kind of "sources" we invite people to add to WeRelate.

So, many web pages are family trees, and those are going to be the some of the most commonly cited. These may not be the "best" possible citations, but they are likely to be the only realistic citation many of us can provide, until another researcher comes along who can actually view some of the sources used on the website. So presumably these too should become miscellaneous?

Ultimately, these sources aren't really going to easily fit into the "author.title." or "place.title" format. I'd suggest "Title. (Publisher/creator of family records (possibly website creator))" or vice versa, perhaps? Another option is to call all family trees Books, but that is pretty artificial.

Ultimately, I think we may want to revisit the decision to make "Manuscript collections" a misc. type, as most of the ones I saw aren't easily going to be titled with "author. title." or by "place. title." They too would be better named as "collection name. Holder of [physical] collection." But for now, misc. will do.

Dallan, perhaps your script for page titles for the misc. pages that do not appear to be Record type should read: "Author. title." if present. If no author, use "title. Publisher." I'm suggesting that users creating misc. source records use the Publisher field and then define the Publisher field to contain either the Book Publisher for books, the website creator/owner for sites, and the physical holder for collections. These seem to be sort of parallel constructs to me, but maybe not self evident to others??.

What your comments, Jilliane and Dallan, don't completely address is the orphaned "finding aids" we are saying are NOT sources, but that we do want to keep access to in some way. Rather than deleting them as sources, we need to find a way to temporarily move them to a category where we turn them into something else - articles, references, finding aids, something. A finding aid may not be a *citable* source for genealogical data, but it is still a research RESOURCE, and whether it is a book, a website, or some other type of document, it is still a useful thing to have access to via WeRelate. So rather than deleting them, we should call them...? Miscellaneous? or something else?

Also, this dicussion also relates to some of the problematic sources formerly called manuscripts. I'll comment on these under that topic. - brenda --kennebec1 15:09, 27 August 2009 (EDT)

How about putting those "Finding Aid" links in "Research Guides" for the particular areas of coverage? Many of them contain only links to the Internet website, and the link could be listed in the particular area research guide. Then the "Source" page for the so-called "Finding Aid" could be deleted. BTW, this idea is #40 in the 101 Ways to use WeRelate.org. --BobC 01:05, 28 August 2009 (EDT)

Moving finding aids onto research guide articles is a great idea! Jillaine has done a lot of work in this area and has created Category:Research guides. Doing this would allow us to keep the links and even list them together on the research guides for the relevant area. After a link had been moved to the appropriate research guide, the Source page could be deleted.--Dallan 17:52, 28 August 2009 (EDT)

Many of these turned out to be for individual family trees, which in most cases (not all) should be listed as MySources since they're relevant to relatively few people. So last year we removed them.
This tells me why I've had to create a number of source pages in recent months for well-known secondary-source websites, while creating Person pages. The catalog includes at least several thousand "family lineage" secondary sources that were published on paper over the past couple hundred years. There are at least a couple hundred on the duplicates list that we've been cleaning up, which tells me the FHL considers them worth having. How is a website that presents the same kind of family history that used to be published in small-run books, any different from a book? Authors publish that kind of thing on the Web now because it's cheaper than underwriting the publication of a book. Even TAG and New England Genealogist give most of their space to footnoted family lineages, not transcripts of original records. You can't say one type of media is okay but another isn't, when they offer the same information. It's the content that counts. --Mike (mksmith) 15:53, 27 August 2009 (EDT)

I agree that there are a number of quality websites for family trees, and that they have family trees that are relevant to a large number of people. By all means create Source page for those websites. The web crawler I wrote found all sorts of genealogy websites (nearly 100,000), most of which were not high-quality, and did not apply to a large number of people. Rather than trying to take the time for each family tree website to determine whether it was worth keeping, which would have taken an incredible amount of time (reviewing nearly 100,000 websites required a lot of time as it was), we had a simple rule: delete family tree websites. I figured that the important ones would get added back.

I believe that if John Doe posts his family tree on a website or creates a few copies of a book at Kinkos, the general rule is that it should be a MySource. Creating a Source page for a family tree should be the exception. Not because of the media, but because of the difficulty of long-term access and the narrow appeal of most family trees. I don't think we want potentially hundreds of thousands of sources added to the Source database that are difficult for others to access (websites get taken down, books with very limited distribution are hard to find) and that appeal to relatively few people. Most family trees that I've seen aren't sourced, so they make poor sources to begin with. This kind of source sounds (at least to me) like the kind of source that should be a MySource, so that when someone searches the Source database they will be more likely to get back sources that they will be able to access and that will appeal to them. That's a judgment call, but I think one that we have to make: Source pages are for sources that will likely appeal to others as well as myself; MySource pages are for sources that will likely appeal only to myself and perhaps a few others.--Dallan 17:52, 28 August 2009 (EDT)

Wouldn't it be relatively simple to just delete those website source pages that have no links to them?

If the collaboration process can actually contribute to an improvement in the quality of sources, getting closer and closer to primary documents, and poorer-quality, more-distant sources are unlinked from as a result, this process could be done repeatedly, eventually removing less useful sources, not by guessing they are not useful, but because there is evidence they are not useful.

Given that the Beta community is small, even one link now will represent a multiplied interest when things are post-Beta. Given the mathematics of genealogy, a website that covers even a few generations will possibly impact many people. For example it is estimated that some of the Mayflower passengers have millions of descendants.

If the real concern is that webpages change and move, it also seems like it would not be too onerous to have a bot that crawls through WeRelate pages and periodically tests any links. Broken links could cause notification of all watchers, and after a grace period to allow motivated watchers to fix the link, then delete it.

Regarding the distinction between items that should be MySource because they "appeal to relatively few people", I am not sure how to quantify that? For example, a limited distribution manuscript (2 copies, both in my possession) that goes back 10 generations and affects thousands of descendants? Or the limited distribution book (2 libraries in worldcat) that covers the 6 generations from my greatX3 grandfather? I am not sure how many descendants are in it since it doesn't exist on my side of the Mississippi, but have been told I am #625, so at least that many. Or the family Bible that lists only one family, maybe 10 people? [P.S. two of these did or do exist as websites, and the third is the type of item I would like to put on the Internet to make it more widely available if copyright allowed, so please consider them as example websites. The point being to get a definition of "relative few people".]

--Jrich 19:15, 28 August 2009 (EDT)

Regarding Source vs. MySource, there's nothing wrong with MySources. MySources serve to store source information that we don't want to return in search results when people search the source wiki. To tell what should go where, the best thing is to put yourself in someone else's shoes: when other people search sources, if this source is something that is likely to be relevant and available to them, then make it a source. Otherwise make it a MySource.

Regarding websites, I really like the suggestion of moving finding aid and other website sources that would be useful for research but probably not for citing to research guides. I think that this makes them more useful than they are as sources, since someone wanting research assistance will be more likely to find them on the guide.

With respect to crawling the sources loooking for broken links, this used to be easier than it is now. Now, instead of giving people 404 errors, a lot of websites have taken to displaying real pages with search links. But the other problem is when websites keep the same URL but change the wording of the page title. I could write a crawler to automatically rename the source pages when this happens, but it seems like a lower priority than fixing usability issues right now.--Dallan 18:51, 29 August 2009 (EDT)

Type:Records [26 August 2009]

I respectfully disagree with this choice. Records would be an okay option for somewhat educated genealogists; but some of our users are not quite there yet. I foresee records being used for family records, Bible records, etc. Face it; many of our users never read the guidelines. I am sorry Mike that this hits your hot button and do I understand the difference in the original records and derivatives but you seem to be the minority in the choice of type. The type as far as know does not appear on your citation or anywhere else. It is merely a classification for grouping of various records. I would appreciate a consensus of opinion before this is set in stone. --Beth 21:26, 25 August 2009 (EDT)

Beth makes a good case (although what alternative would you propose, Beth?). What else might we call this? Official records? jillaine 22:32, 25 August 2009 (EDT)

I would prefer to stay with the original type of government/church records unless someone has an alternative. --Beth 23:10, 25 August 2009 (EDT)

No, I agree, Beth. I kind of like "Official Sources," but that's probably even less meaningful to new users than "Records." In order to distinguish official/primary from non-governmental and derivative, "Government/Church Records" is probably the best choice available. (Though I probably would have separated government from church in the first place -- but they do overlap in various times and places.) --Mike (mksmith) 23:33, 25 August 2009 (EDT)

One of the settled issues Dallan has set and asked us to stop arguing about is that "Records" does NOT separate official and derivative records. It separates geographically oriented collections of records about people and land from sources narrating a larger topic (aka "books"). Under that theory, though, I do agree that removing "gov/church" does not improve clarity.--Amelia 00:21, 26 August 2009 (EDT)

Mmm... so where are we on this topic? Are we then recommending that we leave Government/Church Records alone (as a title)? jillaine 09:50, 26 August 2009 (EDT)

So where does a user put Funeral Home records, Lodge records, Cemetery records, ... --Judy (jlanoux) 10:36, 26 August 2009 (EDT)

"Institutional records"?? jillaine 13:46, 26 August 2009 (EDT)

But that isn't a source type. We're talking source types. There is a field for content and I agree that this is the proper choice for that. Note that word "records" in your choice <g>. Governments and churches aren't the only institutions creating records. I've seen arguments before about whether a particular group of cemetery records should be "Government/Church Records". Making the source type more general avoids this splitting of hairs. --Judy (jlanoux) 14:02, 26 August 2009 (EDT)

As I see it, we either

1. change the source type to "Records" and tell people that they should be creating MySource pages for family or bible records in private possession (which we need to do anyway), or
2. we change it to "Institutional records" and tell people that this includes records created by goverments, churches, and other institutions, or
3. we keep it as "Government/Church records" and tell people that this includes records created by any institution, not just governments/churches, or
4. we change it to "Government/Church/Institutional records" but wow - that's long.

Also, FWIW, if we decide to change the type to "Records", I'll change the label that is displayed on the Source page, but the original label: "Government/Church records" would continue to display in the "History" screen. It would be awhile before I write a bot to update what is displayed on the History screen to just "Records" for all of the source pages of this type.--Dallan 20:58, 26 August 2009 (EDT)

Books that are now online through family search:44,000 plus [26 August 2009]

Dallan, do you have a way to automatically update the pages for books from the FHL that now only have the microfilm number but have been digitized and are online in PDF format. The PDF format is not the greatest. Each page has been entered as a separate entity so you cannot scroll and must readjust the magnification for every page. --Beth 21:38, 25 August 2009 (EDT)

You're joking. Each page? Someone needs to be impeached for that decision. . . . --Mike (mksmith) 23:34, 25 August 2009 (EDT)

It is extremely hard to use and there is essentially no searching. I suspect this was done either to re-use software developed for something else, or with the eventual goal to transcribe the works page by page (small pieces, independently delegatable?). It sure wasn't done with the reader in mind. --Jrich 09:26, 26 August 2009 (EDT)

I think I can get something like this later this Fall. I'm also asking for a list of FHL items that are searchable on their record search beta. I've been told it's possible to get this list.--Dallan 21:01, 26 August 2009 (EDT)

Type: Book [27 August 2009]

I agree with with Mike on this one. The natural inclination is to enter the type book; not records. Will the search engine work for users to find government/church records in books? If so I prefer all books to be listed by author and book title. --Beth 23:55, 25 August 2009 (EDT)

I believe Dallan asked that this discussion stop. As for search, I would imagine that anyone searching for vital/gov records is going to be far more likely to choose a type records (if they choose one) than book, and will be delighted to have the records in books come up.--Amelia 00:10, 26 August 2009 (EDT)

Books that are record compilations that list a single place covered in the title are classified as Records; otherwise they're classified as Books. The search engine won't even let people enter a source type. You'll search on title, author, place-covered, etc. just like you do now. The source drop-down list will continue to list sources by page title, so if you want to link to a book that is a compilation of records listing a single place in the title, you would start by entering the place-covered instead of the author. The reason for this is that the place covered is often easier to determine (and remember) than the first author for these books.--Dallan 10:53, 27 August 2009 (EDT)

Going Forward [28 August 2009]

Everybody is focused on finishing this project and the immediate deadline. But in my daily scan of this page, I can't help but notice a bunch of experienced people who have been working with this project for a long time, asking lots of questions, and I wonder what's a new user going to do a few months from now when this project is over? How many are going to cite a website as a book because they can't find website among the types, or will there be a mass of posts asking how to cite a webpage as a source? What are they going to do when their intuitive search for a webpage title during GEDCOM upload comes back with no source page? Give up? Leave it as a MySource? Create whatever kind of source page they can get the system to create that seems close enough to them? How are you going to explain the philosophy of this new system to users? Add a wizard that asks a bunch of yes/no questions and selects the correct type for them? Or expect to memorize all these bullets describing the types and the rules about each type? --Jrich 09:52, 26 August 2009 (EDT)

Good questions; leaving it is a MySource is very likely. About a month ago, I tested exporting a gedcom to establish if my source citations would transfer correctly. They did not. Until I resolve this problem; I am not using the source system. --Beth 10:14, 26 August 2009 (EDT)

We should test a gedcom upload this after the renaming to see how well it goes, as that will be the most common use by new users-- matching their sources upon upload with the new set of sources. We could (should) change the help-upon-upload text-- that related to matching sources-- that if they have any trouble finding a match, then they should keep it as a MySource. jillaine 10:37, 26 August 2009 (EDT)

These rules can actually be boiled down relatively simply -- just lots of people had questions about unresolved matters and were unfamiliar with the existing rules before we did this. If we don't force people to read all this, it shouldn't be too bad. And if people leave their website citations as MySources, I really don't have a problem with that. If we leave the record set titles as they are (as we should be, but there seems to be some confusion) the book and record matching ought to work fairly well.--Amelia 11:28, 26 August 2009 (EDT)

I left my two cents worth opinion above on the subject of deleting website sources, so I won't reiterate again. I presume by this post that the process is already underway. As I figure about 60-75% of my sources are website based, before I see indiscriminate wholesale deletions of my references to website sources, I'd like to know how much time I have and the recommended method of converting those I feel are worth saving to "MySource" pages. Do I rename? Do I edit the source? Please advise. Thanks.--BobC 00:48, 27 August 2009 (EDT)

This is a good reminder for people doing the deleting: Please don't delete sources that people are actually linking to. Re-classify them to one of the other source types instead.--Dallan 14:08, 27 August 2009 (EDT)

Thank you Dallan for clarifying that. I hope that would also encompass "Watched" source pages by any active user. --BobC 01:20, 28 August 2009 (EDT)

I agree. Don't delete watched source pages either. Re-classify them if you can. If you re-classify them to "should be a MySource", then we'll re-review them later.--Dallan 17:59, 28 August 2009 (EDT)

Speedy Delete - need help [1 September 2009]

I've been trying to keep up with the Speedy Deletes, but they are taking all the time I have and I can't work on duplicates. If anyone has delete authority can help, it would be nice. --Judy (jlanoux) 10:49, 26 August 2009 (EDT)

Judy, I can help with the Speedy Deletes. --Beth 11:25, 26 August 2009 (EDT)

Judy, what are the rules for establishing that a source should be deleted, particularly websites? --Beth 11:32, 26 August 2009 (EDT)

I haven't been questioning the sources in Speedy Delete - just doing it. Several of the people on this committee don't have source delete power and are having to use speedy. In particular, Mike is helping with the duplicates. We gotta delete the extras. You don't have to do the non-sources. I assume the regular speedy people can do those. I just didn't want Solveig to come home from the trip with an overflowing bucket. --Judy (jlanoux) 12:25, 26 August 2009 (EDT)

Thanks. The rule for determining whether a website source should be deleted is:

Is this website something that someone would want to cite on a person/family page?

or

Is this website something someone would want to be referred to in a GenSmarts-like source referral feature?

Websites containing transcriptions of records should be re-classified as "Government/Church records". Websites containing transcriptions of books should be re-classified as "Books" (with the standard caveat about books that are records compilations covering a single place being classifed as records). Websites that are newspapers or periodicals get re-classified as newspaper or periodical. Other websites that people might want to cite on a person/family page can be re-classified as "Miscellaneous". Websites where it doesn't seem like anyone would cite them can be deleted.--Dallan 11:02, 27 August 2009 (EDT)

Finally, don't delete sources that people are linking to (check what links here in the More menu). Re-classify them to Miscellaneous, or should-be-a-mysource instead.--Dallan 14:10, 27 August 2009 (EDT)

Dallan, all the stuff I've sent to Speedy Delete are duplicates, sometimes three or four to the same Source. I've been taking the time to actually merge useful information to one selected page (usually the "Ancestry" page, if there is one) off of the others -- info on reprint editions, etc. The remaining husks are trashed. Most of those have no links; most of the whole catalog has no links, for that matter. But some undoubtedly do, and are already undoubtedly gone. In a couple of places farther up this Talk page, people have commented that deleting a Source page to which someone was linked didn't affect their citation; it only created a red link. And you appeared to agree with that assessment. (CTRL-F for "red link" to see those.) And I'm afraid I don't see the point in reclassifying a duplicate Source to a different Source type. If there are three Source pages to the same, identical Source, and they all have links, you want one changed from "Misc." (which is what they almost always are) to "Book" -- and the other two changed to "MySource"? They're still duplicates. And the clean-up is never going to get done. --Mike (mksmith) 16:08, 27 August 2009 (EDT)

What I'm asking is for people to check if someone is linking to (or watching) the source before they delete it. The vast majority of the time it won't be an issue, because nobody will be linking to or watching the source. But if you ever do find someone linking to or watching a duplicate source that you want to delete, instead of deleting the source, enter #redirect [[Source:title of source to keep]] so that the user will get redirected.

And if you find someone linking to a source that you want to delete not because it's a duplicate but because it should never have been created as a Source in the first place (I've seen a number of those -- sources for individual census pages for example), then re-classify it as a Should-be-a-MySource.

Would that be ok?--Dallan 18:08, 28 August 2009 (EDT)

Okay, no problem with that. (I had read your several "do not delete" posts and thought "Oh, $%&*^*%@") But I went back through the titles waiting in Speedy Delete and didn't find even one with a link, so it shouldn't be too much of a hassle. --Mike (mksmith) 14:12, 29 August 2009 (EDT)

Thank-you. I know there are a lot to delete, but I'm hoping that clicking on what-links-here doesn't take too much extra time, and it could save having to deal with an annoyed user if we delete a source that someone's linking to/watching.--Dallan 18:27, 29 August 2009 (EDT)

Okay, deletion underway. --Beth 12:35, 26 August 2009 (EDT)

Call on me if you need a third person helping. jillaine 13:48, 26 August 2009 (EDT)

Rather than go to this extra work, I'll give temporary admin rights to people who are working on this project so that you can delete the pages yourself. I've just given admin rights to Mksmith. If anyone else is doing a lot of deletions, please let me know below.--Dallan 11:02, 27 August 2009 (EDT)

Just a friendly reminder not to delete a source that has watchers and incoming links. I had to recreate a source yesterday that was deleted when all it needed was an updated URL, which would have taken a few seconds to fix. Instead, I had to start over from scratch. --dayna 10:17, 1 September 2009 (EDT)

Renaming Books? [28 August 2009]

In the duplicate list I am encountering books which someone has renamed to a generic record format. This isn't what we intended, is it? I though books kept their title but we used geography instead of author. --Judy (jlanoux) 10:51, 26 August 2009 (EDT)

You're right. All record sets, in fact, book or not, should be retaining their original title except for censuses and the removal of the place name where it's easy to do so. Otherwise how are people supposed to find their citations? Websites that have changed the original source name of the source might just go on the original page to avoid having more ephemeral web sources, but not books. --Amelia 11:23, 26 August 2009 (EDT)

Census records [10 September 2009]

Why can't a census book keep its title? How is someone going to find it? You are creating duplicates when you do this. --Judy (jlanoux) 12:27, 26 August 2009 (EDT)

Folks could you please point to specific examples? Thanks. jillaine 13:49, 26 August 2009 (EDT)

Sorry, can't because we've got to fix and keep going. On the duplicate list I've encountered cases where an authored work Titled something like "The xxth census of the United States 1850 census of whatever county, state". It was renamed to the generic census form "county, state, united states. 1850 U.S. census population schedule" and will create a duplicate of another source. I finally just had to remove the author and merge the sources because all of the book publication and title information had been destroyed. That source ceased to exist. If someone made a rule that census books are no longer allowed in the catalog, I haven't seen it. --Judy (jlanoux) 14:14, 26 August 2009 (EDT)

I too have encountered this problem with census index sources. Someone added a census webpage index and a FHL census index book together as a combined generic source. I could not add the author of the book since it was a combined source. I did not want to wipe out the source, or choose which link to delete either. I thought I had missed a conversation about this subject. I too do not remember this being decided either. For some reason my previous list is gone the next day so I could not retrieve an example Sorry. Debbie Freeman --DFree 15:14, 26 August 2009 (EDT)

Census books don't keep their titles because they are transcriptions. Census sources are the census schedule. That's it. One per county. This is a combination of the rules Dallan has laid out, and a decision two years ago that we are going to have sources for census on a county level and a national level. Every FHC source and website gets listed as a repository on a page listed as County. 1XXX U.S. Census Population Schedule. and it's former page deleted (or renamed, if that's easier). This eliminates the duplicates, it does not create them. The publication information goes away because that is the publication information for the transcription, which is not useful (both because it's for one transcription only, and because the citation will be most commonly used by people who looked at the films or the images online). Please, please do not delete U.S. Census Population Schedule pages -- I and others have spent untold hours standardizing these sources after the rules were all hashed out long ago.--Amelia 20:08, 26 August 2009 (EDT)

I swore I was through with trying to explain this, but I can't help it. A book of census transcriptions is not the same as the census schedule, and you can't make it true by waving a wand. If you try to take a book created by volunteers in a county society who copied out the censuses for their county and published them, and shoehorn that into the same page as the federal government's official census schedule, you are doing the users of this site a serious disservice. I own exactly such a book, in which the ladies copying down the census added information from the county marriage records and the original land grant records, which ties many families together. It was marketed as Red River County Census, 1850 -- just that. It's an excellent tool. And you've just thrown out all that additional information by making assumptions you ought not to make. I have no problem with a uniform method of naming pages for the actual census. I've created at least 200 of those myself. But you can't call an automobile a four-wheeled bicycle just because you like it that way. WeRelate ought to be encouraging good practices in genealogical research, not sabotaging it by trying to invent its own special brand of source-citation. Frankly, this whole refusal to deal with the world as it really exists is making me a little crazy. --Mike (mksmith) 22:41, 26 August 2009 (EDT)

I never realized that we were renaming census book titles. I disagree. Keep the book title. --Beth 22:54, 26 August 2009 (EDT)

If it's not as simple as look at the title page and copy what you see there, it is going to be done wrong. To quote Dr. John: "If I don't do it, somebody else will." --Jrich 23:04, 26 August 2009 (EDT)

Mike, I do not disagree with you that a book is not the same as a census schedule as far as the SOURCE goes. In your particular case, the book is not just a census transcription. Funny how the rule about only things that are supposed to be the same getting the same page solves the problem. It gets its own page. But as a general rule, the books in the FHL catalog are transcriptions. We could cite them like record sets, but then they are totally inconsistent and harder to find. Moreover, no one is ever going to cite these hard copy books. The existing FHL pages are far more useful to us as the building blocks to create the common population schedule pages that people can eventually use to comment on various resources and point to places to find transcriptions. If we use the fields correctly during the renaming, we will save ourselves a tremendous amount of work.

Let me try a compare and contrast. If we standardize the census record pages, people looking to cite censuses (the vast majority of whom have seen an image of a film these days) will notice that they are almost all the same, be able to predict how to find them and create new ones, and look differently at differently named things like Red Rock County Census. If we don't, people will just pick out whatever looks right when they search the database - be it Smith County Ninth Census of the United States or 1790 Census of Jones County. They won't notice they are actually citing a book or not, and thus their citation will be just as wrong as someone citing the schedule when they've only seen the book (a far less common experience these days). They won't know how to create a source page for a new schedule, and it will be harder to find if it currently exists. The discussion about what transcriptions are best or cheapest or easiest for someone to get will be fragmented across various pages. As I see it, we get no benefits from this system other than the ability to perfectly express the source cited by an infinitesimal fraction of users in a linked page title. To paraphrase Mike, "frankly, this whole refusal to deal with WeRelate source pages as an added-value community resource instead of one person's perfect citation is making me a little crazy."--Amelia 00:20, 27 August 2009 (EDT)

I was going to ask this question myself, so I'm glad that it's been asked already.

Going forward, I think we should allow people who want to, to create separate source pages for census record transcription books, just as for any book. I think that should be our policy. Creating an exception for census books seems arbitrary to me.

For the renaming project: Currently in our source database we have a lot of county-level source pages for census books, and not a lot of county-level source pages for original census records. Most of the census books are now obsolete because digital images of the original census records are available. (In contrast, most other types of records books are not obsolete.) The questions are: how do we create county-level source pages for the censuses, and what do we do with the source pages for the existing census books?

Option (1) is to create original census records source pages titled according to our standard "NNNN U.S. Census Population Schedule" from the existing census book sources, and to list the existing census books as repositories on these original source pages. If there are multiple census books for the same county, we merge them together onto the same source page. If we want, after the renaming is complete I can automatically create source pages for any counties & census-years that didn't get created using this approach based upon when the county was first incorporated (which is listed on the county page). In doing this we'll lose the ability to cite these census books separately from the original census records, because we'll lose the original author+title of the census book. But if someone wants to, they can create a new source page for any census book with the original author and title (still classified as "Records") after the renaming, since our going-forward policy is that books get their own source page.

Option (2) is we keep the census books as separate source pages (still classified as "Records") but retaining the original authors and titles. And after the renaming I'll go through and automatically create county-level source pages for original census records with the standard "NNNN U.S. Census Population Schedule" titles. This will result in many counties having multiple census sources: one for the original records and one or more for the books for that county.

I think choosing between these options boils down to how often we think these county-level census books are going to be cited. If we think that they'll be cited just occasionally, then we should go with option (1), and let people re-create source pages for any census book they want to cite later.

Option (1) makes sense to me, because I believe that the majority of these books are not going to be cited. As an experiment, I will add this rule to the renaming program: if the place-covered of a government/church records or miscellaneous source is a US county and the subject is census records, rename the title to "NNNN U.S. Census Population Schedule". I'll do this later this afternoon and update everyone's lists as well as the duplicates list. We'll then be able to see just how many duplicate sources we'll end up with.

Sound ok?--Dallan 12:37, 27 August 2009 (EDT)

I didn't mean to get quite so anti-book; it was just that the conversation started with people deleting census sources because they didn't look like books and creating new pages for transcriptions and indices and I saw a year's worth of work being derailed. I don't object to people one-off creating new census book sources. If it ever happens, no one is going to notice, frankly. What I object to is having a database full of randomly named sources that are inappropriate for most users of the census, which is what we have now. So I am strongly in favor of Option 1. Option 2 makes things even more confusing.

I love that you can automate the rename. That's beautiful. I don't suppose there's any way that you can either 1) add the general census templates (there's one for each year) or tell me how to run a search so I know which now-properly-named sources don't have the templates (with hand editing, they generally get the renaming, template and category simultaneously)?

Dallan, I've been thinking over the past couple of months about some of these same questions. First, I really don't think the census transcription books are "obsolete." As I noted above, many of them include "value added" info from parallel sources, info that isn't going to show up in the microfilm or on Ancestry. Those books aren't "only" census transcriptions. (Most of the societies that produced these books in past years knew perfectly well that they were in competition with the microfilm, so they made an effort to add something you couldn't get elsewhere.) Having said that, I prefer a system of "every book its own Source" -- which I don't think is going to surprise anyone at this point.
      But I see a good argument, actually, for printed collections of transcribed official records sets (however you define "official") being included on the same Source pages as the original "real" records. The comments I've already seen on some form of this idea have included the word "repository," which confuses me, because we've been using that word to refer to libraries and the FHL and whatever -- repositories where the books may be found. Not referring to a book itself as a "repository." But I was envisioning a Source page for a set of records with a fill-in-the-box section spanning the top-center of the page (which you've said you didn't like to see being left empty anyway), in which one could enter the bibliographical information for a secondary source based on those official records which are the root of the page. There might be one published volume of transcriptions there, or there might be several. On the "primary" census pages, that's where you would put the society-published books of census transcriptions. So a user searching on "1850 census Bedford County Tennessee" would find all the possibilities in one place, on one page, and could cite either the original records or whichever secondary source he had seen and used. (Perhaps the user could select the one he wanted to appear in his particular citation via a radio button.) All those hundreds of volumes of transcriptions of New England town records, for instance, would become closely associated with the underlying originals. Nothing would be replaced, and lost. Nothing would be shoehorned into a container it wasn't meant to occupy.
      I believe this would also mean that all those secondary transcription-type sources could mostly be ignored for now. They could be folded into the Source pages for the underlying primary sources on which they are based a little at a time -- either as people come across them, or as personal projects (I'd like do do that for all of southern Indiana and north Texas, where so many of the Sources I refer to are concentrated), or as long-term volunteer projects. Which means there would be time to devise a properly thought-out method for the merging, a logical series of steps to follow that wouldn't lose important information and wouldn't leave orphaned links, . . . and which wouldn't be under a deadline and cause rises in blood pressure. --Mike (mksmith) 16:49, 27 August 2009 (EDT)

Something is wrong... I'm feeling strange... I agree with Mike. Well, in part ;-) I very much like the proposed source page idea to get us from the benefits of source pages to actual citations. (I don't agree with leaving the current census FHC books around if that's what you're suggesting -- I'd have the ones people use added in when it comes up -- but that's driven mostly by the practical tradeoff of useful pages created en masse v. preserving mostly useless independent pages that are a massive project to fix one by one.)--Amelia 20:01, 27 August 2009 (EDT)

I wish I had better answers, but I'll add a couple of thoughts to the discussion.

The repositories listed on a source page (I hate this use of repositories as well -- we need a better word here), are probably never going to be as convenient to cite as the generic source itself. So we need to be able to generate or human-create a reasonable citation for the source page itself, in addition to a more-complete citation stored for each of the listed repositories. When we have source-matching in gedcom upload for example, most people probably aren't going to go to the trouble of selecting a particular repository. Also, there's extra coding work required to allow people to select a particular repository from the list on a source page, so selecting a repository won't happen right away. And we have the problem of what happens if someone selects a repository from the list, but then someone else removes that repository from the list later. So I'm hoping that selecting individual repositories from a source page will be the exception rather than the rule. I think we need to create enough source pages so most people won't care about selecting a specific repository.

We have to decide now what to do about our existing census transcription books. We can either merge them into source pages with standard titles, create separate source pages for them, or delete them. Due in part to my first comment, I'm in favor of allowing people to create source pages for census transcription books that they want to cite going forward. I just need to know what to do about the ones we have now. The duplicates list show that we don't have too many, usually at most only a couple duplicates per county per census. We just need to decide what to do with these existing sources.--Dallan 19:17, 28 August 2009 (EDT)

Hardly a shock what I think, but I'll write it down anyway. I think doing standardizing now will save a lot of work and confusion later. We seem to be in agreement that few if any people are going to cite the existing FHC sources, and that it's okay to list them on the left side of the 'original' record. So right now, we put in some pain on the duplicates to get the big benefit of getting all the censuses renamed and losing all the titles virtually no one is going to cite. Then the ones that get added later are only the ones that people want to cite for some specific reason, which won't be many. If we delete them, we just have to creat similar records from scratch. If we keep them, they just sit there confusing people. And if it matters, I've merged probably several hundred already, so my approach is both predictable and consistent...--Amelia 20:56, 28 August 2009 (EDT)

Mike, what do you say? I wasn't quite sure based on your last message. If we wantes, I could do the census merges automatically, expand the repository sectiont now to include a cication field, and generate a citation for the original books so we wouldn't lose that info. I could add the census template and other info to the pages at the same time.--Dallan 18:18, 29 August 2009 (EDT)

Well, I'm not sure I follow the latter part of this discussion. You think books (actual books) should have their own source pages (and I agree), and that people ought to be able to create census-book source pages in the future (and I agree), . . . but you're considering deleting the existing separate census-book source pages? Or, you don't think putting transcription collections based on official records on the same page as the original source would actually work (and it might not), . . . but you want to make a single exception for census-books? Sorry, you've lost me. I would say, as I always have, "a book is a book." Let the existing census-transcription books continue to be "books," each with its own Source page, and the whole question of folding in derivative sources (or not) can be postponed until after the present renaming project is finished. I think everyone is too tired to discuss it rationally just now.

Me, I have to go mow the lawn one last time so it doesn't become a savannah while we're away. The 5th-wheel is parked in front, the fridge is being filled, and we'll be outta here in a couple of hours. I'm looking forward to an actual face-to-face discussion of some of these issues sometime this coming week, over barbecue or something. --Mike (mksmith) 10:09, 30 August 2009 (EDT)

This was the question: "We have to decide now what to do about our existing census transcription books. We can either merge them into source pages with standard titles, create separate source pages for them, or delete them. Due in part to my first comment, I'm in favor of allowing people to create source pages for census transcription books that they want to cite going forward. I just need to know what to do about the ones we have now." You seem to be voting for leaving the separate pages for them, merging them later under some to-be-determined protocol. Which is pretty much exactly where we started in this discussion.

We appear to differ on the innate value of the publication information for the census transcription books. I'm willing to sacrifice those details for the greater good, because I don't think it is a loss at all, but even if I thought a handful of people would cite such things, the difference between that citation and a citation to the schedule is so slight it doesn't outweigh the usefulness of getting this all done at once. Mike, as far as I can tell, you're afraid that someone will want that information someday, and that losing it is a symbolic problem inconsistent with good citation practices. I don't think either of us is going to change, and putting this off ignores the not insignificant issue that we have an opportunity to do a tremendous amount of this work in one fell swoop if we rename the existing sources. When we come out of beta, we need to have usable census sources or we'll look silly. I'd rather have a full set of 'official' county schedule sources than a mishmash of 'official' and book sources that should almost never be cited.

Dallan, your call. Everyone else seems to have dropped out of this discussion. --Amelia 19:23, 30 August 2009 (EDT)

I'd like to talk about this with Mike at FGS. For now I'll remove census books from the duplicates list since they'll be merged automatically or we won't title them using this format.--Dallan 10:36, 31 August 2009 (EDT)

Hmmm. Ex parte communication... --Amelia 12:02, 31 August 2009 (EDT)

I had to look that one up :-). I won't make a decision without you. I understand your point of view better than Mike's. I want to understand Mike's better, so I thought a face-to-face discussion would help.--Dallan 18:09, 31 August 2009 (EDT)

What is the decision regarding census books for state censuses? I own a book Alabama Census Returns 1820 and An Abstract of Federal Census of Alabama 1830. I don't want this lumped into the 1820 U.S. Census Population Schedule. There is also another book for Lawrence County, Alabama for the 1820 census and various books for other territorial censuses. --Beth 18:31, 31 August 2009 (EDT)

Regardless of what we do with the existing FHLC books, the going-forward policy is that you can create separate source pages for census transcription books if you want. If you've already created source pages for them, then let me know the titles of the source pages and I'll make sure that they don't get renamed to the standard "NNNN U.S. Census Population Schedule" format in the renaming. You can also check the list of census renames. If they don't appear on this list, then they're not getting renamed to the standard format.--Dallan 18:41, 31 August 2009 (EDT)

Here is an example; I did not create this source.

Source:Henry County, Alabama, records It is not correct to classify this under the 1820 U.S. Census Population Schedule; because the 1820 census of Alabama was not a U. S. Census. Here is another: Source:1820 census, Lawrence County, Alabama Territory--Beth 19:30, 31 August 2009 (EDT)

I did not know that. I thought that the state censuses were on years ending in 5. I hadn't considered that territory censuses would be on the same year. There are 30-40 censuses with the keywords territory or territorial on the census renames list. I'll exclude these censuses from getting renamed to the standard "NNNN U.S. Census Population Schedule" format.--Dallan 20:05, 31 August 2009 (EDT)

What is the final decision on this topic? I also have a question regarding 2 particular sources Source:Alabama early settlers, 1816 : Alabama counties----Mississippi territory and Source:Monroe County, Alabama, 1816 census. In 1816 Alabama was part of the Mississippi Territory, however Monroe County became part of the Alabama Territory in 1817 and part of the state of Alabama in 1819. Also I assume that both of these authors have a valid reason for citing the date as 1816, but these images are on Ancestry and on the third page the date is shown as Mississippi Territory 1815-1817. How are we doing these source pages as records or book or something else? If records do we use the location Mississippi Territory? --Beth 12:35, 8 September 2009 (EDT)

I talked with Curt Witcher (director of the Allen County Public Library) about the question of what to do with census transcription books at FGS. He's in favor of keeping census transcription books as separate source pages, just as with other types of transcription books. A few points came up in the conversation.

• Many people who cite a particular census book want to be able to cite that particular book rather than a generic census source. Many others don't care, but Curt deals with enough people who do care that it's not an insignificant number.
• If we want to match source citations in uploaded GEDCOM's someday, some GEDCOM's will contain links to those books as sources, and we'll want to be able to match them. Having separate source pages for the books will make matching easier.
• We're currently not merging source pages for vital records transcriptions into a single source page with a standardized title, and many of those are going online; are we prepared to merge the vital records transcription book sources in the future?
• We could, and probably should, automatically create source pages using the standard "NNNN U.S. Census Population Schedule" format for all county censuses.
• We may also want to create source pages for originals or digital images of county births, deaths, marriages, land, and probate records using a standard title format (e.g., "county, state, united states. Births", "county, state, United States. Deaths").

After the conversation I'm leaning toward keeping separate source pages for census transcription books. Maybe in the future fewer people will care about exactly which transcription they cite, but for now it sounds like enough people do care that we need to give them the option. We're talking about a few thousand census transcription books total, so I don't think having them will clutter the Source namespace too much. And if we automatically create separate standardized county-level census pages, then people who are looking at original/digital images will be able to cite those sources. Finally, I don't think it would be necessary (worthwhile) to go back and re-do the merging that's already been done by hand.

This also simplifies the answer to Beth's question above: The sources should be changed to type "Government/Church records" so they will get renamed to "Alabama, United States. Alabama early settlers, 1816 : Alabama counties----Mississippi territory" and "Monroe, Alabama, United States. Monroe County, Alabama, 1816 census" respectively.

We've already said that we'd allow people to re-create the source pages going forward if they wanted to, so this question really just comes down to what to do about a few thousand existing source pages. Let's go with a simple majority vote. I already know what Ajcrow, Amelia, Jlanoux, and Mksmith want. What about others?

I vote yes. --Beth 18:51, 8 September 2009 (EDT)

I'm so confused by this time that I don't think I know what "yes" means! But I think I'm persuaded that the book sources of census transcriptions are ok to leave as is. There are so many variations on the "census book" theme; they can't always be easily pigeon-holed into a particular county census page and remain an accurate reflection of their content. So I suppose I'm pretty uncertain about an automated process effectively combining sources in any case.

But I also think that after the renaming effort, a good group of volunteer projects could be generated to review and categorize the "other" census records for a particular area. Some that have no pages linking to them and that are clearly just transcriptions probably could be either integrated into a "standard" county page, or even cross-referenced in the text on the pages. (I have done that with several books that are different editions - referencing the source page for the alternate editions in the text.) --Brenda (kennebec1) 20:52, 8 September 2009 (EDT)

Also, what do people think about automatically creating source pages for censuses (starting when the county was organized), vital records, land, and probate records?--Dallan 15:56, 8 September 2009 (EDT)

I am not sure about this one. We don't wish to create pages for records that don't exist. I do have books that show the censuses that are extant for the different states and counties and books that for the most part are accurate and show record losses for counties. How exactly do you intend to automatically create the source pages? --Beth 18:51, 8 September 2009 (EDT)

I'd be happy to see this auto creation for the census - because we know that 1) counties exist and 2) the census consistently uses counties (when they exist) as a sorting mechanism. But vital records, land and probate records could be handled differently in various parts of the country, in terms of what the fundamental organizational level is. In Maine at least (and I think throughout New England), vital records are never county-wide functions, except incidentally. So making a vital records page for Sagadahoc county would just "mush" together a bunch of otherwise unrelated sources. And in Maine, since 1892, vital records are easiest to group together on the State level. My point is that there is no consistent geographic entity to use in grouping records across both geography (the country as a whole uses the same geography to group these types of records) and time (as long as these types of records have existed).

In Maine, at least, land and probate records are more consistently county-based. But I don't feel confident that that is true for every state. Unless there is a consistent geography sort (country wide), in use over a wide period of time (i.e. all 200+ years of the U.S., or at least since X place began), I don't think we can really create a standard source page. --Brenda (kennebec1) 20:52, 8 September 2009 (EDT)

Well, you know what I think about census records. But on his particular issue, I have to question the generalizability of the experience of a physical library librarian to what our users use. In over two years of gedcom uploads and a year of actively changing hundreds of census links to the standardized format (under what I thought was an agreed-upon format, I changed numerous new uploads to the standard format to demonstrate what it was), I have NEVER seen a citation to a census book. I would say 99% of the ones that specify the source of the schedule specify Ancestry, and the rest don't say. So perhaps there are some people who care (there are obviously a few here) but I think the proportion among people who are still actually visiting a library for census records is going to be much higher than the proportion among people interested and capable of uploading a gedcom here.--Amelia 00:38, 9 September 2009 (EDT)

So, are we debating apples (books) vs. oranges (transcripts) here? Or I am still really confused? Probably the latter... as I'm not a librarian, just a fairly novice user, who continues to find this whole debate pretty mystifying. So what puzzles me is, how do I tell, without the book in front of me, the difference between an actual literal "transcription" ONLY and some other source that is actually a "transcription + value added content" book? Given that all you have to go on in source renaming (generally) is the FHL catalog and/or the description on the WeRelate page, can you tell which is a transcript and should be consolidated and which is a book and shouldn't be?

And, in my run-ins with what I characterize (as an Ancentry user who has become used to having images) as "odd FHL census sources," it seems like there are many with different geography sorts than the County ones that WeRelate has decided to use. There are state-wide sources and locality sources. How would they get renamed or integrated into "standard" census pages? Or, is it reasonable to think that any census source that is NOT in a county geography grouping is not, in fact, a "transcript" but is a "book" with some other content as well?

An example of this is the publication of books in the early 1900s of Heads of Families in the 1790 U.S. Census. Typescripts of the 1790 census, (I thought they were created by the Government?) and apparently sorted into books by state. So is the book "Maine Heads of Families in the 1790 census" integrated into the standard census page for every county in Maine in 1790? (there aren't many...) Or is it added to ALL counties in Maine, because the towns that are today in, say Sagadahoc county, were in 1790 in Lincoln County and in the census? Or does it remain a separate source, because it is a state-wide geography census book? And are these books (which are actually the only other source for the census I've seen, beyond indexes and images on line) "transcripts" or "content-added books?" --Brenda (kennebec1) 22:26, 9 September 2009 (EDT) (who is really lost on this whole discussion at this point...)

Addendum - I guess my questions come from my assumptions in my response to Dallan's questions above. Which are: (1) Assuming no or few pages actually link to these hotly debated FHL-type census sources. (2) assuming they will not, a significant percentage of the time, be obviously or easily integrated in the correct geographic slot. (3) A hope (more than an assumption) that we're not talking about gazillions of sources, but a smaller, more managable number? All of which leads to (4) Ergo, let's wait and deal with them later. But as I don't really know if my assumptions are correct... I figured I'd try to ask to get a better understanding of the problem. --Brenda (kennebec1) 22:40, 9 September 2009 (EDT)

While there aren't many in an absolute sense, there are lots that show up when you search for xxx county census. There are FHC books for at least half the counties I've looked for. My concern is that when people go to try and cite the xxx county census, they will see an FHC book and cite it without realizing that it's meant to be only for people who have seen the book. Why have that confusion, when there aren't going to be cases where people should use the book? But at least if they also get the standard formatted source, they might be prompted to ask what the difference is, and pick the right one.

You're right that sometimes what's there doesn't fit neatly. But a lot, probably the vast majority, does. And what's left can be left in its current state -- this is very much a low-hanging fruit crusade on my part.

How do you tell? I just assume all the census books are transcriptions, frankly, unless the title says otherwise (and it does on occasion). Not a perfect science, but I've been operating under the theory that the benefits of standardization outweigh the extreme unlikelihood of someone caring that I made that call wrong. --Amelia 00:05, 10 September 2009 (EDT)

To answer your question, we're saying that a book gets its own source page. The books you mention will each get their own source pages that will be classified as "government/church records" so the source page titles will be of the form "place. book-title". They will not be combined with the "standard" census pages in the future, which I will create for every county (and state) and that will have a title of the form "place. NNNN U.S. Census Population Schedule". The user will have to choose from among the book and standard census sources, and the (few) people who use the book will hopefully choose the book source, and (majority of) people who use digital images at Ancestry will hopefully use the standard census source. Much of the time only the standard census source will be displayed because we just don't have that many books.--Dallan 15:11, 11 September 2009 (EDT)

It is also not true that we treat other verbatim transcription books separately from the records they transcribe. Where does this come from, aside from the librarians who have recently joined this discussion? It wasn't what we decided on the source committee, and it's not what's posted as the rule in several various places. Change the rule for census if you will, but don't claim that it's going to be consistent with what we do elsewhere. It's not going to be, and it is going to be confusing to the vast majority of users (unless you're going to add "book transcription" after every one of these things, people are going to use them thinking they are the standard format).

If you want to minimize that confusion, you need to create county pages for every county that doesn't have one already, so that they show up in searches and prompt people to understand the difference between the standard format and the other formats. As Beth mentions, this won't be easy to do right, but you've made your bed, so I'm not especially sympathetic!

On the other issue of creating "dummy" pages for citing the "original" records, please don't do that. As was pointed out above, we don't know how those records are held - by county, by year, by state, etc. Many of the collections are probably already here, expressed in the way they are held. Unlike the census, most people at this point do not cite original records, so it won't have the payoff we're looking for either.--Amelia 00:38, 9 September 2009 (EDT)

Ok, I'll automatically generate source pages for census records based upon when the county was organized, but not for other types of records.--Dallan 10:04, 9 September 2009 (EDT)

Dallan, you're clearly wanting more opinions, so I'll give mine, but I've already accepted it won't go this way, which is why I have remained silent.

For general census sources, it has always bothered me that there is no simply-state level source to choose from (i.e., 1850 U.S. Federal Census Schedule - New York). I have never understood why we need more than 50 (plus territories) pages for each year's census. More specific details could be included in the individual citation. In fact, one could argue that we don't need more than ONE page per federal census year. I know I am not alone in using the state level census name for all of my own citations. And without having such a page to link to, I will keep all my contributed census Sources as MySources because I have no intention of going into the thousands of citations I have and converting them one-by-one to the appropriate county-level page.

For the census transcription books done at state or county level, I trust Amelia's experience that there are hardly ANY links to such sources, so why are we bothering?

-- jillaine 10:32, 9 September 2009 (EDT)

Jillaine, I proposed county level originally because we had all these FHC sources that were county level (ironically). If we had them already, I figured, why not keep them. But they should be a standard name so they were easy to cite... Such a mess i involved myself in. Anyway, it's also true that most of the free transcriptions on the web (particularly pre-familysearch pilot) seem to be at the county level because that's a useful bite for people. So I figured that was the best level to take advantage of the 'source tips' aspect of the pages. It also pins down the census pretty narrowly as a citation -- page numbers are such a problem with censuses, but if you know what county someone is in, you can usually tweek the name enough to find it in a search; with common names that's hard statewide. But at any rate, I've come around to the view that having state pages (named in the same format) is a useful thing because of people like yourself that will need it for gedcom linking. There's no reason we can't have both. --Amelia 00:05, 10 September 2009 (EDT)

But at any rate, I've come around to the view that having state pages (named in the same format) is a useful thing because of people like yourself that will need it for gedcom linking. There's no reason we can't have both.

Wow... I'm in love... ;-) jillaine 09:23, 10 September 2009 (EDT)

Ok, I'll add state census sources at the same time.--Dallan 15:11, 11 September 2009 (EDT)

Implementation of Dallan's Plan [2 September 2009]

Dallan, As you may have seen, there are a 'lot' of duplicates this creates. If people don't want to do this work now, I understand, although it would be nice to get at least a jump on it. In the meantime, what should we do to "unduplicate" records where one of the sources is not the population schedule of the census? (agricultural, for instance?)--Amelia 10:42, 28 August 2009 (EDT)

I'm going to remove any census source that includes the words agriculture, agricultural, slave, mortality, or manufacturing in the title. There are a number of books that list both the population and one or more other schedules, but for the renaming I'm going to say that those books will not be renamed to the "NNNN U.S. Census Population Schedule" format.

If there are other titles that shouldn't be renamed, perhaps you could list them here and I'll make exceptions for them.

I'm going to start a new topic on the expected time for finishing the project below. It's grown a lot since it first began, and I'm wondering if we should give ourselves more time to finish up.--Dallan 18:23, 28 August 2009 (EDT)

Hello Dallan, Could we keep the Slave & Native American schedules? WeRelate is one of the few GEDCOM/online that does have African American & Native American Sources. It is a plus for WeRelate. Thanks for considering. Debbie Freeman --DFree 18:21, 30 August 2009 (EDT)

I'm not planning to delete them. I'm just not planning to title books/microfilms that contain other schedules in addition to or instead of the population schedule to the "NNNN U.S. Census Population Schedule" form. They'll be titled using the place-title format. That way they'll appear as separate sources than the population schedule sources.--Dallan 10:27, 31 August 2009 (EDT)

I thought the whole point of using the "NNNN U.S. Census Population Schedule" format was so you could also have "NNNN U.S. Census Agricultural Schedule," "NNNN U.S. Census Slave Schedule," etc. --Mike (mksmith) 21:27, 1 September 2009 (EDT)

It is, it's just that some of these 30-40 census sources that have other-schedule keywords in the title appear to be just for that other schedule, and some have both the population schedule as well as one or more other schedules. I didn't want to try to rename these automatically (too much programming effort for just 30-40 sources), so I thought I'd rename them to place-title format during the automated renaming, and let someone else fix them later (list them under multiple sources, one for each schedule?) if they want. I'll post a list of them now in case someone is interested.--Dallan 14:48, 2 September 2009 (EDT)

Changing to MySources [15 October 2009]

Dallan, I am OK changing some Sources to MySources. The only problem is when sources are saved under MySource it becomes our User name. DFree/MySource as example. Any ideas? Debbie Freeman --DFree 14:59, 27 August 2009 (EDT)

I'm sorry for the confusion. If you find a source that should be a MySource, just change the source type to "Should be a MySource". Then once the renaming process is over, we can contact the authors of these source pages and ask them to rename them, or maybe I'll write something to rename the automatically. But you shouldn't have to rename them now. (If you have renamed a few already, just let me know what they are -- I can back them out.) Thanks!--Dallan 02:54, 28 August 2009 (EDT)

Dallan, during the source renaming project I had followed the directions above by changing some of my personal sources to MySources in the source type block to avoid having them deleted. I just wanted to know whether I should follow through with renaming them individually or if you wrote something to rename them automatically. --BobC 10:56, 14 October 2009 (EDT)

If you wouldn't mind changing them individually I'd appreciate it. I'll be generating a list of all of the "Should be a MySource" sources so that we can contact the authors soon. I'm thinking we'll need to generate the MySource pages manually. (Hopefully there aren't too many.)--Dallan 22:21, 15 October 2009 (EDT)

Source:Sweden. Swedish Kyrka/Church Pictures (http:/freepages.genealogy.rootsweb.ancestry.com/~longstrom/) [28 August 2009]

Hello Dallan & Solveig,

This is on the list, but does not seem to exist as a Source, or link to anything. What would you like to do? Debbie Freeman --DFree 19:07, 27 August 2009 (EDT)

I'd just ignore it. Thanks.--Dallan 18:26, 28 August 2009 (EDT)

Update [29 August 2009]

The review lists should be getting updated in a few minutes with the following changes:

• The bug with the subtitle getting added twice to the title in some cases should be fixed.
• Periodicals are titled using "title (publisher)"
• Miscellaneous sources are titled using the new rule: "use place-title for recordsets that list a single place covered in the title; otherwise use author-title"
• Ancestry-FHLC duplicates should be removed from the duplicates list
• County level US census sources should be titled according using "NNNN U.S. Census Population Schedule"

I need to leave before the process finishes so I won't be able to verify that the changes actually worked, but I'll be back later this evening and I'll check it and fix any bugs then.

To see the changes you'll probably need to force a refresh of the page in your browser.--Dallan 21:42, 27 August 2009 (EDT)

Going to the refreshed Duplicates list, I seem to be finding a bunch of new listings scattered among the sections I've already done. These are mostly "Ancestry.com" listings (one with Ancestry in the page title, plus one or several duplicates), and they're ones that apparently no one has yet touched -- certainly not me. There shouldn't be any new instances of those, should there? Can't figure out where they're coming from, but if the list continues to replenish itself from mysterious sources, I'm not sure how we'll know when we're finished. . . . --Mike (mksmith) 14:25, 29 August 2009 (EDT)

There might be; I'm being more conservative about removing duplicates by adding publication year. I only add publication year to the end of the title now if all duplicates have different publication year.

I could go back to putting publication year to the end of a duplicate title anytime I find one, but I got to wondering if I was removing some sources from the duplicates list that really were duplicates that way.

I'm about to propose that I handle the census duplicates automatically, FWIW.

I'm assuming (hoping) that you're deleting the duplicate sources after merging them, or if they aren't duplicates, then you're changing the title fields in some way so they won't be reaned to the same title anymore. Does that work? If so then we'll be done when the duplicates list is empty. Also, I don't anticipate changing the duplicate list rules anymore.--Dallan 17:52, 29 August 2009 (EDT)

You can see a list of census sources that will be renamed using the "NNNN U.S. Census Population Schedule" title here

The renaming of county-level US census sources to a standard format has added a lot of duplicates to the duplicates list. There are about 1100 sets of duplicates on the list now.

Regarding renaming Miscellaneous sources, in addition to having a records-oriented source type and covering a single place in the title, a Miscellaneous must also have a records-oriented keyword in the title to be renamed using the place-title format. It turned out that a number of our Miscellaneous sources were histories, even though they have a records-oriented source type.

• A 10% sample of sources without a records-oriented keyword is here. These sources will be titled using author-title.
• A 10% sample of sources with a records-oriented keyword is here. The sources will be titled using place-title.

Would someone please review these two lists to make sure that I'm using an appropriate set of keywords? I know that I'm not going to be able to classify every source correctly, but if you notice sources on one list that should be on the other, would you please let me know? I'm unsure about some of the records-oriented keywords that I've used, like "Obituaries". Should obits transcriptions generally be classified as records under our rules, or as books?

Dallan, I looked at the 80% list below. Something you are doing in the sort is pulling some city/business directories into the list. There are also a few place histories left in the list. In particular, the Sources that begin "Ancestry -" are more likely to be on the list incorrectly, although upon reviewing the ancestry ones one more time, most are City directories.

I think the city directories being named by "Place. Title." isn't a bad thing, even tho it doesn't follow our guidelines (extensive discussion above). City directories are certainly "records," albeit not official ones. The ones you rename using the script won't be consistent with others, but they will be findable, I think.

Overall it's a great sort. I didn't get all the way through the list, and I didn't see obituaries catch my eye. I don't have a strong opinion about this, but I'm sure someone will. In general, collections of obituaries won't have one author, though they may have a publisher, and they are likely to be place oriented, so even if it isn't a "correct" representation of records, it is more effective to name them "Place. Title," I would think. -Brenda --kennebec1 08:28, 29 August 2009 (EDT)

PS the handful I saw that were just plain wrong (they were books), I changed to the book type by hand. Mostly they had the word "Gravestone" in their title, but were books about gravestone art and artisans. Maybe not really sources, but good info anyway... and some of the featured graves might well be used by someone in a genealogy. --Brenda. --kennebec1 08:52, 29 August 2009 (EDT)

Of the Miscellaneous source types that have both a records-oriented subject and a records-oriented keyword in the title

• 80% have a single place in the place-covered field 10% sample. These source pages will be titled using place-title.
• 15% list multiple places in the place-covered field, but either one is located within the other, or only one is listed in the title 10% sample. These source pages will also be titled using place-title.
• 5% list multiple places-covered in the title 10% sample. These source pages will be titled using author-title.

--Dallan 18:48, 28 August 2009 (EDT)

Dallan, you might try using "How to" or "finding" or "Guide" as exclusionary key words on the misc. lists. Meaning, if a one place, record oriented title ALSO has one of the key words, it should remain on the "Author. Title" list. Of those words, guide is the one I'm least sure of. It could well be used in the title of a records-oriented source as well as a books source. But on the 10% list, as I continue scrolling through, it seems to be these types of titles (finding aids) that are remaining on the record-oriented one place list in error.

Also, at some point, maybe even after the renaming, you might generate some lists of misc. sources with specific key words (such as these, or even ones with merely individual interest, like I'd be happy to go through a list of Misc. sources with the word "Maine" in them...) for individual review or corrections without time pressure. Then we can address source updatings and corrections topically (eating the elephant in small bites). Scary to admit, but I find it kind of fun to do detective work on different sources... --kennebec1 09:20, 29 August 2009 (EDT)

Thank-you for going through the lists! I'll create an "excludes" list with history, how to, finding, and guide (I'll do a quick duble-check on these words as well). I'll also check into removing gravestone from the records keywords list.

I'll do this Sunday night. Solveig and I are driving back home from dropping off Taylor at college today and tomorrow.

Any other comments, especially with respect to directories or obituaries?

Once the renaming is complete, I'll be happy to generate custom lists for people. I can also show how you can generate your own lists using a back-door into the search engine that lists one result per line, with as many results listed on the page as you want.--Dallan 18:08, 29 August 2009 (EDT)

When will this project end? [29 August 2009]

I know, it's the question we're all thinking. :-)

This project has grown a lot from when it first started. And I think that our Source database will be much better off for it. Originally I was hoping that we would finish the reviews and that I would start the automated renames this weekend so that the sources could be renamed in time for FGS next week. But due to the amount of extra work this project has involved, I'm wondering if we should give ourselves another week (or two).

What do people think?--Dallan 18:56, 28 August 2009 (EDT)

As painful as it can be :-), I would like another week (or two). The lists I'm working on still need help. --kennebec1 08:15, 29 August 2009 (EDT)

I don't know whether anyone else is going, but Judy and I are heading for FGS in Little Rock on Sunday. And I'm a state delegate, so I need to be there early for business meetings. Whether we get anything more done on the lists this coming week depends on (1) whether the KOA in North Little Rock has good (and free) Wi-Fi, (2) what time we get back to the RV in the evenings, and (3) how much energy we have left after a long day of sessions. Dallan, we can both do a little time at the WeRelate table, I think, answering inquiries and handing out literature and all. (Any of the rest of ya'll going to be there?) --Mike (mksmith) 14:07, 29 August 2009 (EDT)

How about if we say that we'll keep going on reviews until sometime the week after next (beginning or end of the week depending upon how things go next week)? That way we'll have the time we need.

Would someone be able to take over working on the duplicates list from Mike next week?

Mike, I really appreciate your offer to help with the table. :-) This would give me a chance to visit the other vendors without leaving the table unmanned. Thabks!--Dallan 17:35, 29 August 2009 (EDT)

I'm working on dups, but not very many at any given time.--Amelia 18:25, 29 August 2009 (EDT)

Can I just say? [2 September 2009]

This is really a great project. I'm deduping sources now, and I'm finding this so educational. I'm learning about all these different types of sources, it's giving me ideas for what else I can research for my own work. It's really quite amazing. And I also love how we're working on this together. jillaine 15:04, 2 September 2009 (EDT)

Will Implementation of source renaming project automatically correct souce citations [9 September 2009]

Just wondering when this project is implemented are the source citations automatically corrected? See the source citation on this page. Person:John Jackson (283) Type was originally misc. now changed to book. --Beth 20:34, 6 September 2009 (EDT)

The source page titles will be corrected and this page will be renamed with the author format. The citation on the John Jackson page will remain as is and will be redirected if someone clicks on it.

Originally, I requested that all of the citations be updated, but I'm having second thoughts. That's a lot of page changes to deal with all at once and you've seen what happens when people see pages getting changed! --Judy (jlanoux) 20:54, 6 September 2009 (EDT)

Yes Judy, I have experienced the reactions; but I would need a notification of the sources that were changed with a redirect to see if my citation needs editing. --Beth 21:15, 6 September 2009 (EDT)

I've already turned off email notifications of changes; it was quickly overwhelming. I did this well before the current project began...

I'd suggest that, if possible, we send ONE email saying that sources are being updated, some will be changed or redirected, and that if you want to receive notifications of every change to a page you are watching, then do XXXX (something). Even if someone has kept their notifications on, the potential amount of email caused by the default notifications of every change will be quickly really irritating and drive away people who haven't been intimately involved in the process.

I know this suggestion is the opposite of the default protocol at we relate and potentially a programming problem, but the human relations problems caused by multiple cascading email notices of changes I think would be far worse. ---Brenda (kennebec1) 08:00, 7 September 2009 (EDT)

This is pretty well a bummer. I don't know how many source citations I have but I guess I will have to manually check each one. I may or may not be watching a source page that I cited. I probably need to redo all of them to make sure that they transfer correctly in a gedcom export so I guess it is not so bad. Just please try and make all of the changes with this project so I don't have to go back and do this again. If WeRelate plans to change additional sources after this project then I will not cite any more sources until it is completed. --Beth 09:28, 7 September 2009 (EDT)

Dallan did indicate that he could probably provide worklists with the citation on pages we watch, similar to the ones we started the project with. I'm going to have to fix as I go anyway, so the list will help me catch them all. --Judy (jlanoux) 10:08, 7 September 2009 (EDT)

How about the following:

• After the sources have been renamed, I'll write a program to update Person/Family pages that cite redirected sources to cite the redirect target. This program will mark any edits it does as "minor".
• Before the redirect, I'll post a message on the watercooler and the news telling people to edit their preferences and turn on "notify of minor edits" if they want to receive these notifications, or to turn it off if they don't. By default "notify of minor edits" is turned off, which I believe is the setting that most people will want -- they won't want to be notified of these source redirection edits.
• At the bottom of the change notification email I'll add a sentence about how we're in the midst of redirecting source citations, and that if they don't want to receive further email notifications of source citation redirects, they should turn off minor edit notifications for the next few days. This sentence would be added to all change notification emails, not just those resulting from source redirections.

Would this approach work?--Dallan 16:08, 8 September 2009 (EDT)

I would think so, sounds like your plan addresses the desire we have to both not drive folks away with an inundation of emails and yet provide info to those who need/want to know. --Brenda (kennebec1) 20:09, 8 September 2009 (EDT)

This works for me too. But I don't know how many people see the Watercooler. Many of the "dump and run" gedcom people now have pages with sources linked because of our merge and clean up activity. Is it against protocol to send a notification email to everyone who watches pages with sources linked? But if "no notice on minor" was the default they probably have it set so maybe it's not a problem. --Judy (jlanoux) 09:55, 9 September 2009 (EDT)

Dallan, if you can "redirect the redirects" auto-magically, that would be terrific! But I would put that third bullet-point at the TOP of the notification email, in large red letters. It's amazing how many people don't bother to scroll their email. --Mike (mksmith) 10:25, 9 September 2009 (EDT)

New approach for duplicates? [15 September 2009]

De-duplicating sources is turning into a very time-consuming effort. It doesn't seem right to me that we're spending so much time on roughly 1000 sources. So I'd like to propose an alternative approach to the problem: For any set of duplicates that all refer to possibly-different editions of the same book, I automatically merge them into a single Source page listing all of the original repositories, with a note in the text of the Source page along the lines of:

This source may refer to multiple editions of the same book. If it is important to you to refer to a specific edition, you may create a separate Source page for that edition with the year of the edition in parentheses after the title.

If we took this approach, then the only duplicates that we would have to deal with would be those that should not automatically be merged in this manner -- say two different books or websites that happen to have the same title. Duplicates that should be merged could be ignored during the review.

What would the people who are doing the source de-dup'ing think about this alternative approach?--Dallan 15:15, 8 September 2009 (EDT)

It is definitely time consuming. As I recall the process was:

1. Open all of them in separate tabs (Firefox)
2. Find true ed info for the "genealogical..." ones
3. Decide which is the 'keeper'
4. Copy necessary data from the others to the keeper (reprint info, Ances, HQ etc); fix pub info if necessary
5. Delete the extras

Will your process keep the publication info from all of the versions? The "genealogical..." ones we talked about handling later. I too have had serious doubts about whether this is worth the time being spent on it. It seems like we'd have to do the first few steps anyway to determine which of your two cases it fit in. Why not just keep what is already cited or has been hand edited and arbitrarily toss one of the extras for all of them? Heresy? --Judy (jlanoux) 16:04, 8 September 2009 (EDT)

I was thinking that it wouldn't keep the publication info from the other versions -- that it would just keep the links to the FHLC, which anyone could follow to find the publication info. So I was thinking that you would just do the first step above (open each into separate tabs), and if it appeared from the information on the source pages that they were probably all for the same book, then you'd close the tabs and move on. The system would keep the first source page and its publication info, and copy the URLs from the other source pages onto it before deleting them.

Alternatively I thought about having the system delete all but the first source, but it seemed like copying the repository URLs for the other sources over to the "keeper" source page, and adding some text about the keeper source page possibly referencing multiple editions, before deleting the other sources would be better than simply deleting the other sources.--Dallan 16:17, 8 September 2009 (EDT)

So we still need humans to review each set of dupes? And the 'truly different' ones will be handled as before? We leave behind the ones which are just different printings for your routine? Do we designate which is the 'keeper' or do you just grab the first one you come to? I'm ok with any of the above if you avoid tossing ones that have been human edited. Some people have done really nice stuff with source pages. The rest are going to be edited by those who care when they cite them. And those who don't care...well they won't care.--Judy (jlanoux) 17:27, 8 September 2009 (EDT)

Oh that would be lovely if you could do it automatically. I haven't seen a single useful source among several dozen merges I've done, so I really can't get fussed about perhaps screwing up the distinction between editions and reprints. If you do keep one set of sources and their fields, keep the Ancestry ones. Nearly all the FHC ones have the wrong publication fields because they use the filming information instead of the book information.--Amelia 00:43, 9 September 2009 (EDT)

Yes, we'd still need humans to review each set of dupes. And the truly different ones would need to be handled as before -- change the fields on the source page (e.g., source title) so they won't get renamed to the same page title. But I'm hoping that you won't have to do anything with the vast majority because they'll be for the same book.

I'll select as the "keeper" the one that has been human-edited or linked-to (if any), or the Ancestry one (if any), or else the first one if none of them have been human-edited/linked-to or are from Ancestry.--Dallan 09:38, 9 September 2009 (EDT)

That works for me. Once I get caught up with home duties, I can start back on the dupes later today. --Judy (jlanoux) 09:42, 9 September 2009 (EDT)

Has this been done? Or will it be? Are we just supposed to be reviewing and leaving the ones that you can handle, or are you going to run your protocol and we do what's left?--Amelia 00:32, 14 September 2009 (EDT)

Just keep reviewing and leave the ones that are ok to merge if you would. Once people have finished reviewing the duplicates list, I'll merge the remaining duplicates.--Dallan 15:29, 15 September 2009 (EDT)

Website sources [15 September 2009]

BobC and I have been slowly working our way through the website sources:other subjects lists, and although many could be deleted AS SOURCES, they would be useful in Research Guides or as links on a surname page. I am concerned that there is no way we can get through even this one list, never mind the Finding Aids or records oriented websites lists, before the source renaming happens.

Is there some way to either 1) neither rename or delete these pages until later or 2) rename them in such a way as to be able to "pick out" these former-websites-now-miscellaneous-or-something else sources later? For example, could they all temporarily have their subject field renamed "Website?"

I suspect the records-oriented websites with a place name will be fine being renames to the "place. name." format and don't need any more (or less) review than any of the other autogenerated sources. But the other lists of websites are more problematic, and unless we plan to hold off on renaming for another month or so, I doubt we can get through them in time. It is quite laborious to research and move them... --Brenda (kennebec1) 10:06, 12 September 2009 (EDT)

We had agreement that the websites could wait until after the rename. It is the duplicate lists that are holding things up. We need to devote our attention to them. After the rename, we can pitch in to help deal with the website list. --Judy (jlanoux) 10:46, 12 September 2009 (EDT)

Sorry, Judy, Brenda and I must have missed that "agreement;" who all signed on to that? I thought the purpose of the whole list was to take care of those under our names and then "adopt" the other ones that needed work. Yes, there was a lot of discussion about duplicates, and I figured that was being handled okay, so I saw a need to work on these "various other lists" before the Speedy Deleters axed those pages (and thence those referenced links). I play better when there's less of a crowd on the playground anyway. So you and the others can finish repairing the merry-go-round, and Brenda and I will work the swing set, okay? --BobC 13:33, 12 September 2009 (EDT)

In the "Website sources question" topic I touched on the issue of deferral, right before you stepped in. I repeated it in response to your concerns and Dallan repeated his concurrence. Sorry you missed it, we were trying to allay your concern. I just stepped in here in response to Brenda's concerns that you couldn't get done in time. There isn't a deadline for websites. It is the duplicates that are holding up the project. The duplicates must be resolved before rename. Cleaning up duplicates is very laborious and does require that extras be deleted. I'm not sure why you have a problem with that - they are extra copies. --Judy (jlanoux) 14:05, 12 September 2009 (EDT)

Appreciate the clarification. --BobC 23:02, 12 September 2009 (EDT)

Brenda,

I created a "tree" called "Convert to Research Guide" and added each of such resources to that. It enables me to go back later and convert them. I don't know if there's a way you can add those to MY tree, but that would be ideal. Otherwise, create your own, and I can work from that. jillaine 10:33, 12 September 2009 (EDT)

Thanks, Jillaine. How do we access it? --BobC 13:42, 12 September 2009 (EDT)

Jillaine, (is that your actual first name? it is beautiful and unusual... 2 syllables?) would it work if you "shared" the tree you created (i.e. emailed it to BobC and I from manage your tree)? I'm not all that knowledgeable about how the technical aspects of werelate work, and I vaguely recall that there are limits to the shared tree (like, I'm not sure if it updates "my" version of your tree if you make changes on "your" version of your tree?) If that were true, that would be a fatal flaw.

I confess to making "on the fly" the other day a Category:Finding Aid for pages that should become finding aids (generally translate to "become part of a research guide or surname page." That would be another way to make a master list to deal with later.

Actually that may work better than the tree conversion. And quicker (as an intermediate fix). I like it. Good call. --BobC 23:12, 12 September 2009 (EDT)

Yes, the agreement to defer websites (not delete them, not rename them, just leave them alone) wasn't clear. And looking at the list, I'm honestly not sure which of the 4 website lists we want to leave alone and manually handle, and which of the lists we want to try to rename automatically?

• record-oriented subject and one or more places-covered - should we rename these automatically using place-title format?

I think they probably can be renamed using the place-title format, awaiting further review. Assuming that the type on these records won't change, we'll still want to go through them later.--Brenda (kennebec1) 23:00, 13 September 2009 (EDT)

Sounds good to me.--Dallan 15:32, 15 September 2009 (EDT)

• record-oriented subject but no place-covered - can we handle these manually?

Not that many of these; wait and handle manually is my vote. --Brenda (kennebec1) 23:00, 13 September 2009 (EDT)

• Finding Aid subject - handle these manually as well?
• other subject - handle these manually too?

yest to both, --Brenda (kennebec1) 23:00, 13 September 2009 (EDT)

You could add some of these to a tree, and you can view someone else's trees by going to User page and clicking on the "view" link next to the tree you want to view, but it's problematic: You can't add or remove pages from another user's tree. This is a general problem that I need to handle in a better way -- allowing people to really share trees where both people can add and remove pages from the tree. But that won't happen until early next year.

Would it work if I just continued to refresh the lists that we decided to handle manually each morning after the rename? They could continue to list sources of type "Website", and if you delete a source or change its type, then it would disappear the next morning.

What do others think about this approach?

that works fine for those that remain websites. There are some I have changed to something else (misc.) in order to keep them (earlier in the process), but could ultimately be moved to finding aids and research guides etc, or are in a grey area of maybe sources, maybe research guides.... But I have kept a list on a User Page of the ones I want to rereview later, so that will help. Really, we will still have to review many of the misc. manually over time to give them a more accurate type. But that doesn't affect the renaming for now. --Brenda (kennebec1) 23:00, 13 September 2009 (EDT)

P.S., Judy is right that the only thing holding up the source renaming right now is the duplicates list. I recently changed the approach (the above topic: New approach for duplicates), which should make it go faster, but there are still several hundred left to check. --Dallan 17:44, 12 September 2009 (EDT)

Finding aids [14 September 2009]

Can whoever is doing finding aids handle these from the duplicate list? They are inventories, not records. --Judy (jlanoux) 12:13, 12 September 2009 (EDT)

Meurthe-Et-Moselle (France). Archives Départementales. Inventaire-Sommaire des Archives Départementales Antérieures à 1790, Meurthe-Et-Moselle,

   * Inventaire-sommaire des archives départementales antérieures à 1790, Meurthe-et-Moselle, archives ecclésiastiques, série G
   * Inventaire-sommaire des archives départementales antérieures à 1790, Meurthe-et-Moselle, archives ecclésiastiques, série H 

Meurthe-Et-Moselle (France). Archives Départementales. Inventaire-Sommaire des Archives Départementales Antérieures à 1790, Meurthe-Et-Moselle, Série

   * Inventaire-sommaire des archives départementales antérieures à 1790, Meurthe-et-Moselle, série D
   * Inventaire-sommaire des archives départementales antérieures à 1790, Meurthe-et-Moselle, série E supplément
   * Inventaire-sommaire des archives départementales antérieures à 1790, Meurthe-et-Moselle, série 7F (collection de Mahuet)

I am going through all of these Inventaire-sommaire records. --Brenda (kennebec1) 15:10, 12 September 2009 (EDT) ...eliminating duplicates in the manner modeled by someone? where inventories from the same locale with the same title are combined on one page. For now, I am making them (as my model did), Government/Church records so they are titled by locale, which will make them easier to sort out and is accurate (they are inventories of official records of various sorts). But ultimately, these pages (many fewer for each locale) and/or these links should be referenced on the research guides. Boy, there are alot of them! --Brenda (kennebec1) 09:53, 13 September 2009 (EDT)

Thanks for handling that. I haven't explored finding aids yet and it was difficult to try and switch gears when I was doing duplicates. --Judy (jlanoux) 20:19, 14 September 2009 (EDT)

Another Ancestry cleanup run? [15 September 2009]

Dallan: Will you be able to do another Ancestry run? If so, I'm done with my duplicate list. All of the lists will need to be reviewed again once the dups are done because of changes being made. But I'm done for now. Can't handle anymore. --Judy (jlanoux) 17:03, 13 September 2009 (EDT)

I'm not sure what you mean; the duplicate list is updated every morning. If you've finished your duplicate list, how about calling it "good enough" rather than going through the list yet again? I'm sure a few mistakes will be / have been made, but I wonder if we'll soon approach the point of diminishing returns (and I'd like to put the project behind us :-).--Dallan 15:38, 15 September 2009 (EDT)

Ok, I see you noted that you'll merge what we leave. That will do. I'm very diminished by this project! --Judy (jlanoux) 16:53, 15 September 2009 (EDT)

Beth needs some help/leading article/FHL film number [14 September 2009]

I need clarification on 2 items.

• Will this automatic operation remove the leading article from the title?

• • Will this automatic operation remove the FHL film number from the title?

--Beth 18:59, 14 September 2009 (EDT)

Re leading article - The rename will do this.

I have a lot of duplicates on my list caused because one was with and one without the "The". This is evidence that the rename is going to make that change.

Re film number - If the film number is in the Title box on the page, we have to take it out (I did have a few of these). If it is just in the page title and not in the title box, the rename will take care of it. --Judy (jlanoux) 19:26, 14 September 2009 (EDT)

Thanks. --Beth 19:36, 14 September 2009 (EDT)

Interment.net [23 September 2009]

In tackling the duplicates, I have a website question. There are 61 WeRelate sources that come up when I search for Interment.net. Some are state-wide subsets (Source:Cemetery Records of Oklahoma), some are county subsets (Source:Cemeteries of Middlesex County) and some are cemetery specific (Source:Zion Hill Cemetery).

My sense is that Interment.net should be a repository, and that (somehow) all the cemeteries covered by Interment.net should be places and/or sources, with interment.net cited as a repository. For today, I will move the Interment.net source to a repository and leave the other Interment sources (all of which are website sources and can be dealt with later) as sources. This came up, tho, because of a duplicate source name. The Interment.net sources are mostly named as variations on "Cemetery records of XXXX," which is likely to cause dupes. Should I rename them "Cemetery records of XXXX (Interment.net)"?

And, in the duplicate I'm looking at, Oklahoma Cemeteries, the *other* "Cemetery records of Oklahoma" is a FHC microfilm of a particular typescript list of some Oklahoma cemeteries and individuals interred. There is no indication that the FHC list has any similarity or overlap with Interment.net list (although the FHC list does not provide any specifics on which cemeteries are covered) so I don't think Interment.net works as a repository for this source. Any thoughts on this? --Brenda (kennebec1) 09:58, 15 September 2009 (EDT)

I should note that on a few of the Interment.net sources, Jbean has already begun renaming to add (interment.net) at the end of the title, so making that change would be consistent with previous changes. And it doesn't *appear* that any of these are linked to yet. --Brenda (kennebec1) 10:03, 15 September 2009 (EDT)

Hi Brenda, this would have been my next question. I quit because I felt as if I was hampering y'all's great work because; I obviously am not current on all of the decisions and the project page is not detailed enough to discern all of these decisions. I hate to butt in again; but interment.net should be the source; at least as indicated by Mills in Evidence Explained. Page 233: ...You will also note that your source for these online records is the database-not the cemetery or the marker.--Beth 11:07, 15 September 2009 (EDT)

I suspected this would have been your next question as soon as I looked at it... :-). I like the solution offered below and I don't think anyone has used any of the internment.net sources. --Brenda (kennebec1) 16:05, 15 September 2009 (EDT)

I vote for deleting the internment.net state sources. This is an artificial division of the internment.net database, which doesn't really distinguish between states any more than it does any other way. If we want ONE page, that's fine, but state-specific pages are silly and useless.--Amelia 12:34, 15 September 2009 (EDT)

Agreed. --Beth 13:19, 15 September 2009 (EDT)

I think it would be appropriate for the Internment.net site to be identified as a Repository. The state listings are nothing more than subcategories of the website, therefore should not be included separately. The individual cemetery is a Placename and the page identifying an individual posting for a decedent should be considered a MySource if used within WR. That's my take. --BobC 12:18. 21 September 2009 (EDT)

Well, Bob I have been mulling over your comments all day and fail to find the logic in classifying Interment.net as a repository. First of all it does not meet the criteria presently listed for a MySource. Second, it does not meet the criteria presently listed for a repository. The definition of a repository is an archive, government office, library, or other facility where research materials are held according to page 828 in Evidence Explained. If we are actually trying to educate people on the correct method to use in citing their sources; why would one wish to muddy the issue by classifying Interment.net or Findagrave as a repository?--Beth 23:55, 21 September 2009 (EDT)

Thanks for taking the time to mull, Beth, but I have to agree with Brenda & Amelia. With regards to the well-respected Evidence Explained, let's go beyond the book and take a look at both ends of the spectrum of the word Repository, both from a broad non-genealogical use of the word and from specific uses of the term within WeRelate. According to Dictionary.com, the Random House Dictionary definition of a repository is, "(1) a receptacle or place where things are deposited, stored, or offered for sale. (2) an abundant source or supply; storehouse: a repository of information;" the American Heritage® Dictionary of the English Language defines the word as, "A place where things may be put for safekeeping;" and according to About.com, a repository is, "a collection of resources that can be accessed to retrieve information. Repositories often consist of several databases tied together by a common search engine."

Since those definitions may not be sufficiently grounded in genealogy for you, let's take a more specific look at WeRelate references to a Repository. According to the WeRelate Repository Portal it says, "Repository pages are for physical locations of genealogical material such as archives and libraries, and for significant virtual locations such as major genealogy websites." I would say both Internment.net and Find-A-Grave are both significant virtual locations and major genealogy websites.

While I concur that Inernment.net and Find-A-Grave are not physical locations or cannot be considered as original depositories for these records, there is precedence for citing major genealogy websites as repositories here at WeRelate, such as the discussion at an image talk page in identifying Footnote.com as a repository, discussion at a source page talk page in changing the listing of the US National Archives from a Source to a Repository, and the discussion at a source pages talk page regarding the designation of Google Books and NEHGS as repositories.

If you read my message above a little closer, please note I did not say that the Internment.net site should be listed as a MySource, only the individual page for a decendent's listing, if used by a user as a source for their ancestor or family member's data. We had previously decided that an individual newspaper article or obituary should not be shown as a Source, but should be identified as a MySource instead. See Help:Source pages and its talk page for further information and discussion on the topic.

To me, this is a very logical approach, despite Mills's fine work. This is not "muddying the issue," but emphasizing standardization, consistency, and uniformity within the WeRelate community. --BobC 06:10, 22 September 2009 (EDT)

I can see both sides of this issue. The original thinking behind having repositories for major genealogical websites was that websites like Ancestry.com and NEHGS contain multiple collections, and each collection would be considered a Source in the website repository. A website like Interment.net contains a single collection. I think we ought to have a Source for it, so that people who want to can cite it directly from a Person/Family page by including information about the specific record they're citing in the source citation without having to create a MySource. Since it is a major genealogical website it also qualifies as a repository. I don't think it hurts to have both, although I don't think we want to have too many website repositories.--Dallan 10:49, 22 September 2009 (EDT)

This type of issue is what led me to request that you defer handling of websites until after the rename when we can discuss it properly. There was no need to try to rush into what is going to become a complex (and probably touchy) issue. There are some websites that obviously should be publicly-maintained sources (source namespace). There some other websites that obviously should be privately-maintained sources (MySource namespace). Figuring out where the divider is in the middle is hard! Writing rules for a user to follow is harder! Getting users to follow the rules is probably impossible!! --Judy (jlanoux) 11:15, 22 September 2009 (EDT)

The reason this came up was not because Interment.net was a website, but because there were numerous sources for various subsets of the Interment.net site, and they were causing duplicate page names. Interesting that we are now debating this, as based on the earlier responses and to resolve the duplicates, I have already deleted all the subsidiary page names, unless they were cited by someone. I believe only one was actually used as a source, and that person did in fact use Interment.net as the source (the whole site, not a subsidiary), adding the details in the person page. I redirected that source to the generic Source:Interment.net Cemetery Records Online.

And by my mistake, there is actually both a Interment.net Respository and an Interment.net Source at this time, so for now, the status quo is that Interment.net is both a repository and a source.

As Judy notes, getting users to follow our (hotly debated) rules is probably impossible. And in this case, I agree with Dallan that it probably doesn't matter a great deal. With Source:Interment.net Cemetery Records Online, a user can add details as they need to the person page. If a user chooses to keep MySources, Repository:Interment.Net allows others to find the specific reference if they need it.

I do think variations on this discussion are going to continue to come up, as we don't have a good method or place to put the wide variety of finding aids on the internet. Some are sources for info, in part. Many are pages of links. Some are really good pages of links, like Cyndi's List. They are neither repositories or sources, but they can be useful places for people to begin searching. When such a page is specific to a topic (i.e. immigration), it can be linked to in the research guide. When a page is specific to a place, it can be linked to on the Place page. When a page is specific to a surname, it can be linked to on the Surname page. But for other, less significant portals (i.e. most of those smaller than Ancestry), how to categorize them and put them somewhere in WeRelate so they can be useful to searchers is a dilemma. --Brenda (kennebec1) 11:46, 22 September 2009 (EDT)

I am not debating the issue of deleting the subsidiary pages; I agreed with that decision. I just happen to disagree that Interment.net or similar databases should be listed as a repository as opposed to a source page that one may cite. I agree that Ancestry and Footnote are repositories. If you wish to have a source page and repository page it doesn't bother me in the least as long as I am not forced to use the repository page. I am interested in attempting to follow the current standards in genealogy. Three genie programs that I know of use Mill's as a standard, Legacy, Roots Magic, and Family Tree Maker. I see no valid reason for WeRelate to do any differently. It would seem to me that following some sort of standard would make the site more user friendly and enable easier transfer of data by gedcom export. Also sometimes in the future if WeRelate were to have its own genie program; it would be helpful if the program was compatible to other well know genie programs. I am more concerned with my source citations than which repository I obtained the information from. --Beth 12:35, 22 September 2009 (EDT)

Probably good we're having this discussion (debate) now. Although it can be a fine line between the two (source and repository), I think it can be pretty simple when looked at rationally rather than dynamically: the repository is where the source was obtained, the source is a record of what was obtained, and the data is the information that the source contained. In the case of an obituary record, the respository is where the document was located, the source is the obituary itself, and the data elements (such as names, dates and locations) are what the source identified. So the way I see it, the repository for these obituary records can be the local courthouse, the Internment.net site, the local newspaper, or my personal collection. And the source will always be the obituary or an extracted record of it. I know this view would probably not comply with the Elizabeth Shown Mills standard (or Beth's). --BobC 14:59, 22 September 2009 (EDT)

I think it's good to have these "what's the right way to" discussions to a point because it helps us clarify our thinking around possible community policies. But I think it's also worthwhile to ask: (1) is this something that I can clearly explain to new users? (2) Is it necessary for them to obey this policy in order to participate effectively in the community environment? And (3) If they do it wrong, am I willing to police the pages and correct them? Whenever we tell people that not only do they have to learn a new system to use WeRelate, but they also have to change they way they record their genealogy, we lose some people. So I think we have to be careful about the policies we set. We have to pick our battles. In cases where it doesn't matter greatly which way we go, or even if we go both ways, I think we can leave it at that.

To me this seems like a battle that's not worth it. We could recommend one way or another on a help page for repositories, or list the reasoning behind each approach and let people choose what suits them.--Dallan 15:48, 22 September 2009 (EDT)

Since we seem to be getting somewhat off-topic (relating to the Source Renaming Project) with unclear concensus and uncertain interest to go any further on this here, I'll not beat this horse to death. In an effort to answer those questions you raised, Dallan, I'd like to put some information together on the Repository Portal or it's talk page to see if we can formulate a more consise practice for users (new and old) and promote an acceptable guideline on both the use and labeling of sources and repositories, taking in mind professional publications like Beth emphasized above on the subject and other genealogical programs' standard practices in those areas. --BobC 21:48, 22 September 2009 (EDT)

Hi Bob, sounds good to me or we can move the discussion to your talk page and then regroup on the portal talk; either one.--Beth 22:05, 22 September 2009 (EDT)

Good idea, Beth. I've moved the discussion over to the Repository Portal Talk Page for all who want to join in it. I considered putting it on my User Talk Page as you suggested, but I don't want it to appear as I am trying to claim ownership of it. I'd like the debate, analysis, recommendations and guidelines to be viewed as a community discussion rather than a private one. Thanks. --BobC 21:48, 22 September 2009 (EDT)

The duplicates are declared done - ready for merge [21 September 2009]

--Judy (jlanoux) 12:41, 17 September 2009 (EDT)

I still see a lot on the dupes page. (Including some A-F that I don't remember seeing before.) Are we sure it's done? jillaine 13:09, 17 September 2009 (EDT)

Dallan can merge the rest. I still don't know how new things keep appearing. I'm starting to wonder if we're just undoing each other's work. But we've certainly taken care of the "big rocks". I vote we quit. --Judy (jlanoux) 13:52, 17 September 2009 (EDT)

Wonderful! I've decided to restructure French places to have a shorter place hierarchy before renaming sources so I'll do that for the next couple of days, and then I'll start renaming sources on Monday. Terrific job everyone! I'd like to echo Jillaine's sentiments in the Can I just say? topic. I love how we worked on this together.--Dallan 20:19, 17 September 2009 (EDT)

I'm going to run back through my original "review" list, as many of the changes I've made since then have shown up there, and I'll be able to see if I've created any new duplicates in my Inventaire sommaire project -- which I think I did. Dallan, is that ok if I still make a handful of changes in the next day? --Brenda (kennebec1) 20:49, 17 September 2009 (EDT)

In the french sources I see on my list, some words are being inappropriately capitalized, I think, like "et" and "es" and maybe "du."--Brenda (kennebec1) 20:58, 17 September 2009 (EDT)

Thanks. I've added et, es, and du to the do-not-capitalize list. Du will be capitalized if it is entered by the user as capitalized, which is also how we'll handle words like Des and La - capitalize them only if the user enters them capitalized.--Dallan 11:01, 21 September 2009 (EDT)

Dallan, When do you plan to do the duplicate merge? We will probably have to check afterward to make sure they all were handled. I'm really ready for the rename to be over, I've got life issues piling up. --Judy (jlanoux) 00:21, 18 September 2009 (EDT)

Over the weekend the system removed the "region" level from the titles of about 36,000 places in France and changed Quebec to Québec in another 1500 places. It also merged about 3600 sets of duplicate sources. About 140 sets didn't get merged because people have edited/linked to more than one source in the list. I figured Solveig and I would review those 140 sets today and do a combination manual/automated merge to finish them up. I've updated the duplicates list so you can see what's left if you're interested.

The other thing for today is to update sources that refer to French or Quebec places with the renamed place titles. I'll also modify the online add-source function to use the new title format.

Dallan, please hold on this. It looks like whatever the system did affected the German place names, too-- and incorrectly. See my message to you on your talk page. I'd hate for you to change all the sources that point to German place names if those place names aren't correct yet. jillaine 12:05, 21 September 2009 (EDT)

Thank you for being so careful to notice these! I looked at the places that you listed, and they're not incorrect. They're all redirects that eventually, after following multiple redirects, point to the correct place. The changes I made yesterday were simply to remove text after the bar (|) in the redirect line; I didn't change where they pointed to. The program that I plan to run today that updates the places listed on source pages will update the places to have the final redirected title. I'll review the list of sources being updated before I run it to double-check for German mistakes, but I'm pretty sure we're ok. Once we're done with the source renaming I plan to write a program to delete un-linked-to place & source redirects, and to update the redirects that are linked-to to point to the final redirect.--Dallan 15:48, 21 September 2009 (EDT)

Hopefully once that's all in place the source renaming can start tomorrow (yea!). I estimate that it will take approximately two weeks to rename the 900,000+ sources. I don't want to rename too quickly because I don't want to slow down normal use of the system.--Dallan 11:01, 21 September 2009 (EDT)

Re: the 140 left, It looks like the repository merge didn't take care of the Ancestry books. For some reason these got left behind when you did the first Ancestry cleanup - usually because there were 3 or more copies. I think I got them all down to 2 now so you should be able to do the Ancestry run again and cleanup the rest. They don't have people linked.

This is great news! I give it a triple yea (Yea! Yea! Yea!) --Judy (jlanoux) 11:19, 21 September 2009 (EDT)

The main thing happening here is that the FHLC source in the Ancestry-FHLC pair has either been linked-to by someone or edited by someone. I purposely didn't merge those sources automatically. Solveig is going through them and copying text comments from the FHLC source onto the Ancestry source (which there hardly ever are, but she did find one FHLC source so far that someone had edited and added a comment to). Then she either redirects the FHLC source to the Ancestry source if someone links to it, or deletes the FHLC source if no one links to it. And I copy over the FHLC repository information to the Ancestry source automatically as usual.

In addition, there were a handful of sources that really weren't duplicates but were going to get the same title because the first part of the title was exactly the same. I shortened the title or author in those cases so that the differing part of the title would be included in the page title.

It is great news. We're almost there :-) --Dallan 15:48, 21 September 2009 (EDT)

UPDATE: I just created the list of sources that link to redirected place pages. There are 86,000 sources, mostly French parish registers. I hadn't anticipated there being so many French sources. It's going to take a day to update them, and I want to finish updating before starting on the renaming. So the renaming will begin Wednesday, not Tuesday.--Dallan 17:04, 21 September 2009 (EDT)

"Other Subjects" List [21 September 2009]

(Discussion regarding other subject list transferred from project page to this talk page)

Some of these should be changed to "Repositories" rather than remain as "Sources" --BobC 14:32, 27 August 2009 (EDT)

Please consider many of these pages as potential "Research Guide" articles and references rather than outright deletions. --BobC 14:34, 27 August 2009 (EDT)

As time permits I will attempt to transfer propspective Finding Aid links to appropriate area Research Guides so clean-up crew can delete those pages if appropriate. --BobC 09:01, 28 August 2009 (EDT)

I'll help; I'd like to see many of these pages available as background sources, even if not directly genealogical sources, on a variety of topics. --Brenda (kennebec1) 16:17, 31 August 2009 (EDT)

As of this morning, I've gone through and transferred all of the WR source pages under this grouping identified as Native American related with external links and added them to the Native American Research Guide. Links transferred, I nominated those source pages for the "Speedy Delete" process. --BobC 11:34, 1 September 2009 (EDT)

Right now am working on all of the WR source pages under this grouping identified as African American related with external links and am adding them to the African American Research Guide. As links are transferred, I am tagging those source pages for the "Speedy Delete" process as well. --BobC 13:44, 1 September 2009 (EDT)

Am also renaming library, museum, historical & genealogical society source pages to "Respository" pages instead, verifying web links in the process, and adding repository contact information. --BobC 13:46, 1 September 2009 (EDT)

Can a source page be renamed into a different type, or do you have to recreate them to make them a repository?--Brenda (kennebec1) 10:18, 9 September 2009 (EDT)

Yes, I don't know what magic Dallan worked, but just rename it and overwrite Source: with Repository:. Why are we talking on the project page? --Judy (jlanoux) 10:43, 9 September 2009 (EDT)

Brenda (and others interested in helping), I could use assistance on the surname-type and placename-type pages in transferring external links to appropriate Research Guides. Then those pages can be identified for "Speedy Delete." (I have tried to be very careful not to add the "Speedy Delete" function to any page that is being watched or has WR links.) It's a slow process, but hopefully worth it rather than seeing these source pages (and accompanying links) disappearing under impatient hands. --BobC 13:55, 1 September 2009 (EDT)

Thank you Bob! I've given you temporary admin rights during this project so you can delete the pages yourself once you're through with them. That should speed things up a bit.--Dallan 16:32, 1 September 2009 (EDT)

Thanks, Dallan ... I think :) --BobC 16:39, 1 September 2009 (EDT)

Back to the outside level again: I'll work on moving family trees pages to surname pages as links on those pages. Do you think that is the best way to preserve those sites? i.e. is a Surname page the equivalent of a "Resource guide" for a name? That's what makes sense to me. --Brenda (kennebec1) 10:18, 9 September 2009 (EDT)

I agree that the Surname Page is probably the best place to record those family tree or surname links. If nothing further is linked to those source pages and if no one is watching them, then they can be tagged for {{Speedy Delete}}. The two page types (Surname page & Resource Guide) could be considered equivalent: the Surname page holds general information on and links for surnames from unconnected family trees, and Resource Guides contain data about and links for sources of genealogical information on places or groups. Thanks again for the assistance. --BobC 10:37, 9 September 2009 (EDT)

Brenda, does it appear to you that this page is twice as long as it used to be? I was going to work on it a little before bed but got intimidated by the size of it. Maybe tomorrow while I'm sitting at a couple airports... --BobC 00:03, 12 September 2009 (EDT)

I have noticed that the lists do change, but my approach has been to not even scroll all the way to the bottom at first glance, 'cause if I did I'd never start :-). So I can't say for sure. My own "sources to review" list has definitely grown; I think many pages I edited turned up on it if I didn't turn off watch this page. Also, I suspect that as others rename and change pages (i.e. change the "type"?), they may well end up on the sort that creates this list. And I've really only barely started...--Brenda (kennebec1) 07:52, 12 September 2009 (EDT)

Which list are you talking about? Your own personal page will grow - it contains pages where you're the last editor, regardless of whether you are watching the page. But the website lists shouldn't be growing, unless someone is changing the subject from "Finding aid" to "Other", in which case the pages will move off of the "finding aids" list and onto the "other" list.--Dallan 20:12, 17 September 2009 (EDT)

I was referring to the other subject list. I haven't noticed page histories showing a change from "Finding Aid" to "Other," but it probably doesn't matter, as I would just as soon take care of them correctly rather than having them tagged for "Speedy Delete" without having the links transferred to where they can be utilized better by WR users. --BobC 13:26, 21 September 2009 (EDT)

I fixed a bug where the subject was being listed on the line above where it should have been, but otherwise I can't figure out how the list could be growing. The Other list currently has 1230 sources, the Finding Aid list currently has 6010 sources, and the "Records" lists have 2045 sources (at least one place listed) and 95 sources (no place listed). We can monitor the sizes from now on. I wish that I had kept track of the sizes of these lists previously, but I didn't.

I have completed the Jewish-related links, either renaming invalid sources as repositories (for societies and libraries), or deleting unused source pages and transferring all external links to the Jewish Research Guide. --BobC 12:43, 21 September 2009 (EDT)

Congratulations!--Dallan 16:02, 21 September 2009 (EDT)

Moratorium on adding source pages? [23 September 2009]

Someone mentioned on Solveig's talk page the possibility of having a moratorium on creating new sources during the renaming. I hadn't planned on doing this, although I suppose I could. I hadn't planned on doing it because starting tomorrow newly-added sources will follow the new title standards. But maybe I'm missing something. What do others think?--Dallan 15:52, 22 September 2009 (EDT)

Well, . . . I've taken time out the past few days from volunteering to actually import and tidy up a GEDCOM. :) Which includes creating a few new sources (and a lot more MySources, for grave markers). But I've been following the new style! --Mike (mksmith) 11:32, 23 September 2009 (EDT)

Final Rules Please? [15 October 2009]

Dallan,

Could you please post the final final final rules that drove your renaming and that we can now incorporate onto the Help:Source page titles page? A few of us are having different understandings of the final rules, so it would help to receive The Final Word on the subject. Thanks.

jillaine 10:20, 1 October 2009 (EDT)

I just commented on the Help talk:Source page titles page. Overall I think it looks excellent! And yes, the author-title format does include given and middle names in the title.--Dallan 13:58, 2 October 2009 (EDT)

Dallan, thanks for the response; however, your comment over on Help talk:Source page titles was specific to only two items: author name (and I think you meant Surname, Given, not Given and middle) and geographic. It would be helpful to see the final final for all aspects. Is it the same as the final proposal above? Sorry to ask this of you, but I think we really need to know what actually drove the renaming on all counts so that have The Final Word to refer to when there is disagreement about what we each remember. Thanks. -- jillaine 07:47, 4 October 2009 (EDT)

Author name can include surname, given and middle. The latest version of Help:Source page titles includes all the rules that I can think of. Are there additional things that it needs to address?--Dallan 22:32, 15 October 2009 (EDT)

Status report please! [15 October 2009]

What's the status of the source renaming project? Is the auto-renaming complete yet? I ask because I'm still seeing old-style page titles in search results. -- jillaine 13:16, 4 October 2009 (EDT)

I've been taking a peek at Admin, log to see where we are. It's on the W's so I hope we're close to done and not that there's another round pending. The system has been exhibiting the most bizarre behavior the past few days - I want it over. I don't need this much excitement in my life.

It's amazing how slow "blinding electronic speed" can seem. --Judy (jlanoux) 14:01, 4 October 2009 (EDT)

Is it over? Is it over? Logs show most recent agent renaming of sources was on those Chinese character names. Status please?! jillaine 16:23, 8 October 2009 (EDT)

Yes, it's all over finally -- nearly 1,000,000 sources have been renamed. There are about 100 sources left that I need to rename or merge manually [3]. I'll do this tomorrow, then I'll start removing the old source page titles. But I'm going to delete them at a slower pace.--Dallan 22:42, 15 October 2009 (EDT)

Watch for trouble [15 October 2009]

Congratulations Dallan for getting the renaming past us.

I've been discovering some problems (well we expected that, didn't we) that I want to warn others to watch for. I think most of the source users were watching this page.

One thing that I just came across is a case where a source was renamed several times by us working the list. Then the project run came along and renamed it to one of the interim names. As a result the link from my Person pages just went to a redirect page and stopped dead because of circular references. It's easily fixed by removing all the redirects and relinking my sources and notifying the other user. But we want to be alert for this type of thing.

I don't know if it's possible to run a list to see if a redirect string doesn't ultimately go to a real source page.

It would be nice to have that list of citations we talked about so we could start checking them. --Judy (jlanoux) 14:14, 6 October 2009 (EDT)

Other trouble, none too huge:

• Vital records reviewed before the final decision was made (which follow one of many other interim rules) - or just misc. titles where the system guessed wrong.
• Double names - the system doesn't remove the place name from the title itself, so there are some very long duplicative titles.
• Funky authors - I have a few I found that I don't know why weren't on my list, but they are ones where I entered the author a long time ago as, e.g. "Ed. David McBride", which is now in the title as is.--Amelia 14:33, 6 October 2009 (EDT)

Once I've deleted un-referenced redirects (the old source page titles), I'll write a program to update the redirects that are being linked-to to point to the final redirect target. I'll look for source pages that don't redirect to real source pages at the same time. Then I'll write a program to update Person and Family pages to link to the final redirect targets as well. So the redirects will be taken care of automatically, but you'll have to correct any source page titles that are messed up manually I'm afraid.

Can you remind me what "that list of citations that we talked about" was?--Dallan 23:08, 15 October 2009 (EDT)

Re-including place in records sources [26 December 2009]

Beth has pointed out that when we excluded the place from the source title field of a "church/government records" source, the place doesn't show up in source citations generated by her desktop genealogy program. For example, the citation for Source:Jefferson, Alabama, United States. 1850 U.S. Census Population Schedule that is automatically generated by Legacy is simply 1850 U.S. Census Population Schedule because the source doesn't list an author and we removed "Jefferson, Alabama, United States" from the source title field.

According to ESM, the citation for this source ought to start out along the lines ofAlabama, Jefferson County. 1850 U.S. Census Population Schedule.

I know there are a lot of questions about how to best generate citations, so let's start with an easy one:

• So that desktop genealogy programs like Legacy can properly generate citations from WeRelate-exported sources, should I automatically update "government/church records" sources to include the place covered as the author if there is no other author?

For example, should I update Source:Jefferson, Alabama, United States. 1850 U.S. Census Population Schedule to have an author of Alabama, Jefferson County?

Doing this would allow programs like Legacy to generate proper citations from WeRelate-exported gedcoms. Any thoughts?--Dallan 20:37, 28 November 2009 (EST)

Well, first of all, I think you guys are confusing Elizabeth's format for a footnote with her format for a bibliographic entry (Evidence, p. 74). A "citation" would be a footnote. And the format for that is so far from what WeRelate exports, I don't think it's worth arguing about. The thing is, there are dozens of genealogy desktop programs out there, not to mention Ancestry and so on, and they're all going to construe a GEDCOM import differently, so you're not likely to come up with an export from this site that's going to make all of them happy.

Beyond which, I really don't think it's reasonable to expect any software, or website, to produce a block of text with associated footnotes that could be sent to a journal for publication without editing -- if that's what you're after -- so I don't think I understand the point. The format for sources I use in TMG have all the necessary information, but they don't really follow Elizabeth's style, either. And so what? They're for my convenience, and if I'm writing an article I'm going to re-cast everything into a better, "public" format anyway. TMG isn't (realistically) capable of producing a finished, grammatical, readable text that follows all the style rules for every publisher, either. No software is. No website is. Don't worry about it. --Mike 09:17, 29 November 2009 (EST)

Mike, the point is that when one exports a gedcom; the location of the census is not in the gedcom. While I personally probably will never use the gedcom export feature, I believe that the state and county information should be in the export. --Beth 09:37, 29 November 2009 (EST)

Is the state and county information included anywhere in the source when it is exported? -- Amy 10:20, 29 November 2009 (EST)

Yes, Amy. The information is included in the hyperlink to the source page. The hyperlink is in the comments field of the source. However, if you include the hyperlink in your program generated reports, you have formatting for the Wiki in the source page title. You would need to edit every church/government source exported. --Beth 10:39, 29 November 2009 (EST)

Isn't there a better way to make sure that state and county information is included in the export besides "faking" an author? I would be far more in favor of changing the title fields back, since it's really more title we're talking about.

Until this is fixed, though, we probably should suspend the rule about removing duplication in a title (i.e. the title for the California Death Index is just Death Index 1940-1997 so the page title says California, United States. Death Index 1940-1997).--Amelia 10:32, 29 November 2009 (EST)

I haven't retested export since the rename. So I created an export file and tried it in Legacy and in TMG. RootsMagic won't open the gedcom so I couldn't see what happened there. I see the problem. Prior to the rename, we had the place as part of the Title field on census records. As part of the rename the place names were stripped from the title. So it does look like they need to be added back for the export. Neither program has a field for place in their census template so they must assume it is in the title box. The information does appear in the comment type box in both programs where you have put the link to the page name.

Now we have an issue of where to put the place. As Mike pointed out, the format for footnote and for bibliography are different. I don't see a place for these different styles in Legacy. TMG has three output fields for Full footnote, short footnote and bibliography. But since there's no place name field, I'm not sure how they intend to automatically implement it. While developers keep trying to do it, the "build a citation" approach always introduces more complexity than it solves. I just type them in skipping all the separate fields. TMG provides this capability. I don't know Legacy well enough to see what it does. I tried adding a new source in RootsMagic to see what they do. They do have a place for the jurisdiction. I didn't explore to see what they do with it. Bottom line is that different programs do different things. Putting the whole deal in the title field is probably the safest approach. But I prefer to use the United States version for the census and put the place in the detail part of the citation. So this wasn't a problem for me, except that I don't care for the Legacy handling of this. I never saw the point in adding hundreds of extra sources to my list. Maybe because I'm lazy and typing is very difficult for me.--Judy (jlanoux) 10:39, 29 November 2009 (EST)

I agree that it would be best to have the county and state in the GEDCOM export; however, I don't think that putting it in the author field is the best way. Appending it to the title field would seem to be the best/safest place to put it. -- Amy 10:48, 29 November 2009 (EST)

I agree with Amy that it should be in the Title field, not author. Also, I noticed that you export the link to the history page as "author" and that could be a problem. It makes sense for Person pages to do that, but could cause confusion on Sources between the author of the source and the author of the source page.--Judy (jlanoux) 10:55, 29 November 2009 (EST)

Ok, I'll add to my todo list automatically re-instating the "place covered" in the title field for "government/church records" sources. While not perfect, this should result in a reasonable bibliographic entry in most desktop genealogy programs. And I'll change the "add source" functionality to do the same thing for new sources. As for the possible confusion with author, I don't know what else to call that link, so I'll leave it alone unless someone has a better idea. Thank-you for the feedback!--Dallan 01:00, 26 December 2009 (EST)