Jump to content


Photo

Adventures in Extreme Splitting


  • Please log in to reply
66 replies to this topic

#21 Kevin

Kevin

    Advanced Member

  • Members
  • PipPipPip
  • 31 posts

Posted 23 April 2013 - 04:11 PM

Ancestral Quest actually has this feature, allowing the same citation (not a copy and paste) to be used as a source for multiple facts. This also has the advantage of all the references to the citation using the Same footnote number which avoids duplication and clearly indicates they are all based on the same source.

#22 Nettie

Nettie

    Advanced Member

  • Members
  • PipPipPip
  • 1647 posts

Posted 26 April 2013 - 04:07 PM

Jerry: I like your last message , IT was a great relief when the Place List changes came to be that there was not any more global changes except from the Master List.. Clean up is much easier now. :)

In the Source listings, I have it in my mind and keep it there that Source Citation Main screen [mine is yellow] I do not know change but can change the Details as that is for the individual. I like this and works for me.

Genealogy:
"I work on genealogy only on days that end in "Y"." [Grin!!!]
from www.GenealogyDaily.com.
"Documentation....The hardest part of genealogy"
"Genealogy is like Hide & Seek: They Hide & I Seek!"
" Genealogists: People helping people.....that's what it's all about!"
from http://www.rootsweb....nry/gentags.htm
Using FO and RM since FO2.0 


#23 Bgart

Bgart

    New Member

  • Members
  • Pip
  • 1 posts

Posted 11 May 2013 - 11:03 AM

I am brand new to RM6 after having used FTM 2012. I have a basic problem with source citing that I believe relates to lumping and splitting as described here, but I'm not completely sure. The scenario is that I have multiple sources for a birth. One being the vital record at the state level, and another being the birth certificate from the hospital.


Each of the certificates has the name of the father, but the State level birth certificate only has an age for the father, whereas the Hospitals certificate has the actual Birth Date. As far as I can tell right now. RM6 wants to treat the Birth fact as a combined event, which includes the Birth Place, Birth Date etc. I would really like to enter the source once, it is an individual record after all, and use it multiple times for each fact, not only the Birth date of the subject of the Birth certificates, but also as a source for the name, birthplace and birthdate of the Father and Mother.

It doesn't seem that I can do this as easily in RM6 as I did in FTM. Since I have the birth date for the father on the Hospital Certificate, but not the state level certificate, I would like to use it as a source for the Birth Date, Birth Place, and Father's name. Likewise I would like to use the State Birth certificate for Name of the father, Birth Place, but not the birth date. Since births are compound events, its harder to do in RM6 than other programs. At first blush I should create a birth fact for each source, and only fill in the details that are included in the matching source. So in the case of the Birth Ceretificate from the State, I would create a Birth Fact for the Father where its a source for the Name and Place but for Birth date, I would provide a range, not a specific date. As far as the master source and details. They don't change a particular source. Only its use in supporting a fact. I'm sure I'm missing something here but it looks to me like RM6 should be providing better granularity in saying what specific part of a fact is supported by a particular source.

What I would rather do though is create the source once, and link it to each subpart of a fact with one click. If the source doesn't support that support its not added to that subpart. FTM lets me do this, but is a pain for actually entering the source itself.

Hmm for some reason the formatting is not working for me and the text looks too small. Sorry


#24 Jerry Bryan

Jerry Bryan

    Advanced Member

  • Members
  • PipPipPip
  • 3589 posts

Posted 11 May 2013 - 06:37 PM

Since I have the birth date for the father on the Hospital Certificate, but not the state level certificate, I would like to use it as a source for the Birth Date, Birth Place, and Father's name. Likewise I would like to use the State Birth certificate for Name of the father, Birth Place, but not the birth date. Since births are compound events, its harder to do in RM6 than other programs. At first blush I should create a birth fact for each source, and only fill in the details that are included in the matching source. So in the case of the Birth Ceretificate from the State, I would create a Birth Fact for the Father where its a source for the Name and Place but for Birth date, I would provide a range, not a specific date. As far as the master source and details. They don't change a particular source. Only its use in supporting a fact. I'm sure I'm missing something here but it looks to me like RM6 should be providing better granularity in saying what specific part of a fact is supported by a particular source.


You are correct that the connection of a particular source to to a partiular fact in RM does not have the granularity for which you are wishing. RM does handle births as compound events for sourcing purposes, and the same is true for deaths or burials or marriages or any other fact. Any source citation for a fact may only be connected to the fact as a whole and may not be directed specifically to the date or to the place or to the note, etc.

By contrast, sources themselves may be as granular or as ungranular is you wish. You may treat a book as a whole as a source, for example, or you may treat a single page of a book as a source. And even if you treat a book as a whole as a source (the most common way to do it), within RM itself you can still treat a single page as a source detail. And whichever way you do it within RM, the way the citation appears in a report can look exactly the same no matter how you organize you sources and source details within RM.

The topic of this thread is about the advantages of making source citations be extremely granular. But whether a particular citation of a source is extremely granular or extremely ungranular or somewhere in between, the citation of the source may only be applied to a fact as a whole. And of course the citation of a source, granular or not, may be applied to one or more facts for one or people if so indicated.

Jerry

#25 Nettie

Nettie

    Advanced Member

  • Members
  • PipPipPip
  • 1647 posts

Posted 12 May 2013 - 06:52 AM

Jerry answered this really well. :)

One basic reason that I switched many years ago from FTM to FO/RM was that FO/RM allowed us to have one fact/event and many sources attached and each source could be used for many different fact/events without making a new copy. When I started we were limited to under 5 sources per fact and now many more :) is possible. Also each source that a particular piece of the puzzle/fact/event could be identified. In RM a birth fact/event has the date, place and [I may have up to 10 sources per fact/event] source/s with notes/research notes/comments sections.

For me, having two birth facts with each a different source, would be confusing in a Family Group Sheet report or Narrative report. One birth fact with multiple sources, each source with notes as to what is in the source, is a better solution. :)

Genealogy:
"I work on genealogy only on days that end in "Y"." [Grin!!!]
from www.GenealogyDaily.com.
"Documentation....The hardest part of genealogy"
"Genealogy is like Hide & Seek: They Hide & I Seek!"
" Genealogists: People helping people.....that's what it's all about!"
from http://www.rootsweb....nry/gentags.htm
Using FO and RM since FO2.0 


#26 zhangrau

zhangrau

    Advanced Member

  • Members
  • PipPipPip
  • 1518 posts

Posted 13 May 2013 - 06:37 PM

I agree with Nettie on this:
-
I use the Source Details, Source Research Notes, and Fact Notes to describe which information (or partial information) came from each source. The Fact Notes are included in the Narrative reports, but the Source Research Notes are not.

#27 Nettie

Nettie

    Advanced Member

  • Members
  • PipPipPip
  • 1647 posts

Posted 13 May 2013 - 07:34 PM

I agree with Nettie on this:
-
I use the Source Details, Source Research Notes, and Fact Notes to describe which information (or partial information) came from each source. The Fact Notes are included in the Narrative reports, but the Source Research Notes are not.


Yes the Source Research Notes and Comments can be part of the Narrative Report. :)Check the choices under Source Settings when setting up the Narrative Report. If you have the same information in the Fact Notes and repeated in the Source Research Notes then don't check that choice. They will print twice... :o

Genealogy:
"I work on genealogy only on days that end in "Y"." [Grin!!!]
from www.GenealogyDaily.com.
"Documentation....The hardest part of genealogy"
"Genealogy is like Hide & Seek: They Hide & I Seek!"
" Genealogists: People helping people.....that's what it's all about!"
from http://www.rootsweb....nry/gentags.htm
Using FO and RM since FO2.0 


#28 Jerry Bryan

Jerry Bryan

    Advanced Member

  • Members
  • PipPipPip
  • 3589 posts

Posted 19 May 2013 - 10:44 AM

For the first time since embarking on my Extreme Source Splitting project, I'm processing a death certificate. I decided to spend some extra time to "get this one right" on the theory that however I handle this particular death certificate will be the same way I handle numerous death certificates in the future. Plus, I may wish to go back and reprocess death certificates I already have in my database to conform to my new process.

It seems to me that there are three main things that I have to "get right". First is how I name my Master Sources. Nobody but me will ever see these Master Source names, but even so having a logical naming convention should be very helpful to my research. What seems to be working the best is something like the following:

*death certificate: Doe, John Earl
  • The * is just to be sure all my new style Master Source names sort in front of all my old style Master Source names. If I were starting a whole new RM database I would not include the *.
  • The "death certificate" without the quotes part of the name assures that all Master Sources for death certificates sort together under List->Sources.
  • The : is to separate the "death certificate" identifier from the person's name. Making the Master Source name something like "*death certificate, Doe, John Earl" without the quotes is like finger nails on a blackboard to me because a comma is being used to separate the type of Master Source from the person's and also is being used to separate the person's last name from his first name. It's just a horrible and out of balance lack of parallelism.
  • I put the last name before the first name to get the names to sort by last name in List->Source.
I'm not including the death certificate number, the state, the county, the year of death, the repository, or anything like that in the Master Source name. It seems totally unnecessary, and the appropriate information will appear instead in the footnote sentence.

The second main thing that I have to get right is the storage and naming of the death certificate image files. Instead of storing them by family or by geography, I'm storing all of them in one flat folder structure where the name of the folder is deathcertificates. This procedure will create a one-to-one relationship between my Master Sources for death certificates and my images for death certificates, so that they will mirror each other perfectly.
  • Within the deathcertificates folder, I'm naming the files something like Doe_John_Earl_death_cert.tif.
  • Death certificates are usually only one image, but in the event that I ever have multiple images for a death certificate then I would use a single tif file that was a multiple image tif file.
  • In the rare case that I might have more than one John Earl Doe in my database, I would add birth and death years to the file name and also to the Master Source name.
  • I'm linking the Master Source for the death certificate to the image. In accord with the principle of Extreme Splitting, all citations that use the Master Source will therefore inherit the image.
  • I'm adding the image to the Media Gallery as a file rather than as an image to enable the image to be displayed by an external viewer rather than by RM itself. That's because an external viewer is a much better viewer than RM, and because the image window for an external viewer is independent of RM rather than being locked into an RM sub-window from which you cannot escape without losing your view of the image.
The third main thing I have to get right is the footnote sentence, and this is the biggie. I feel like a deer in a headlight trying to figure out the proper footnote sentence.
  • I use RM's free form source template, but that doesn't prevent me from peeking at RM's built-in source templates for ideas. RM has two source templates for death certificates, one for local death records and one for statewide death records. Both of these templates are based on Evidence rather than on Evidence Explained. I've misplaced my copy of Evidence, and I find trying to use the built-in templates for death certificates to be very confusing. If I try to fill in the required data, I don't know what to fill in plus I end up with the "comma comma comma" syndrome where data that I omit because I don't know what to enter results in ugly ",,," things in my footnote sentence.
  • I can put my hands on my copy of Evidence Explained. The most relevant pages seem to be p.430 and p.467. The latter page gives the following as one of several examples of footnote sentences for death certificates.

Oregon Department of Human Services, death certificate no. 79-006405, Tzilla Lilah Miller (1979); Center for Health Statistics, Portland.

  • I research mostly in Tennessee. Death Certificates are filed with State of Tennessee, State Department of Health, Division of Vital Statistics. But there is a 52 year privacy rule on death certificates in Tennessee, and for the most part you cannot get death certificates at all from the Division of Vital Statistics. Instead, you can only get death certificate that are older than 52 years old and you can only get them from Tennessee State Library and Archives (better known as TSLA), a totally different organization than the Division of Vital Statistics. I assume that after 52 years, Tennessee death certificates are sent from the Division of Vital Statistics to TSLA.
  • I used to get death certificates by sending a check to TSLA. What you had to know to get such death certificates was the person's name and date of death.
  • After a while, microfilm for the death certificates started showing up in several local libraries that I use. The libraries simply purchased the microfilm from TSLA. What you had to know to get death certificates from the microfilm was the reel number and the death certificate number. Which reel number it was depended on the county and the death certificate number, and you had to look up the death certificate number on a different reel of microfilm that was only an index. The clerks at the TSLA had to follow the same procedure to find death certificates for you when you mailed them a check with the person's name and date of death.
  • It is now the case that images for the exact same microfilm are available and indexed at ancestry.com. You lookup death certificates by the person's name. There is no way to tell from ancestry.com what the reel number is, and death certificate numbers are not indexed. The only way to tell what a death certificate number is at ancestry is by looking at the image itself after finding the image by the person's name.
  • Under these circumstances, I've decided to go with something like the following for the footnote sentence.

Death certificate #18852, John Doe (1933), ancestry.com (Tennessee, Death Records, 1908-1958), viewed 18 May 2013.

  • Note that I'm citing the person's name in the footnote sentence exactly the same way as appears on the death certificate, including those cases where it says Mrs. John Doe rather than her maiden name of Jane Elizabeth Smith. That information plus the year of death should always be sufficient for anyone to find the certificate at ancestry.com. If somebody wants to find the death certificate the old way at TSLA or on microfilm at a library, the person's name as it appears on the certificate plus the certificate number plus the year of death should be sufficient.
  • Actually, there's one minor glitch in my scheme if you want to find a death certificate on microfilm. From 1914 to somewhere in the mid-1920's, Tennessee death certificates were filed at the Tennessee Division of Vital Statistics but nevertheless were filed by county. That meant that for Anderson County there was a death certificate #1 and a death certificate #2, etc. For Bledsoe County there was a death certificate #1 and a death certificate #2, etc. The same procedure was followed for each county. Therefore, the microfilm for those early years is also organized by county. So for those early years, the death certificate number is not sufficient to find the certificate on microfilm without knowing the county. (This is obviously not an issue at ancestry.com) So my footnote sentence may need to include the county). Perhaps something like the following might be better for those early years. The trouble is, if all you have access to is ancestry.com you will be entirely unaware of this issue.

Death certificate #18852 (Knox County, Tennessee), John Doe (1933), ancestry.com (Tennessee, Death Records, 1908-1958), viewed 18 May 2013.

  • The name "Tennessee, Death Records, 1908-1958" is precisely the name that ancestry.com gives their database of Tennessee death certificate images. It's a very useful thing to know because it's very easy to search a particular database at ancestry.com provided you know the name of that database, and searching in that manner has huge benefits in eliminating false matches from your search.
  • Most of my death certificates are from Tennessee, but I also have a goodly number from states such as Kentucky, Missouri, and Texas, plus a scattering from other states such as California and Florida. In my experience, the processing of death certificates is not identical among the states, but it is very similar. So I think my basic scheme should work fairly well for most states. Of course, for those cases where I have to write off for death certificates, the format of the footnote sentence will have to be changed a little to adapt to the fact that I didn't obtain the death certificate online.
What's a little frustrating is that my footnote sentences for death certificates look so different from the example footnote sentences from Evidence Explained and also look so different from the footnote sentences produced by RM's source templates that are based on Evidence. It's like, who am I to think I know better how to make footnote sentences than the experts? But my footnote sentences for death certificates make sense to me. And I can't seem to put the information I have available to me into either the Evidence Explained format or into the RM source template format. It's like putting a square peg into a round hole.

Jerry

#29 Jerry Bryan

Jerry Bryan

    Advanced Member

  • Members
  • PipPipPip
  • 3589 posts

Posted 19 May 2013 - 11:07 AM

Just a little follow-up and clarification on Tennessee Death Certificates, TSLA, and ancestry.com.

When I do a search at ancestry.com where images are involved, I usually just go straight from the search results page to the image I want. From the image page, there is little citation information that is available. However, you can also go to a "View Record" screen which has a bit of a transcription of the record plus some source information. For the example I've been using (a real person, renamed as John Doe as an alias), the following appears at ancestry.com on the "View Record" screen.

Source Citation: Tennessee State Library and Archives; Nashville, Tennessee; Tennessee Death Records, 1908-1959; Roll #: 8.
Source Information:

Posted ImageAncestry.com. Tennessee, Death Records, 1908-1958 [database on-line]. Provo, UT, USA: Ancestry.com Operations, Inc., 2011.

It's not clear to me how ancestry.com intends for a user to combine the Source Citation information and the Source Information information into a proper citation. The Source Citation data makes no mention of ancestry.com, for example. And I know from experience that the TSLA roll numbers do not correspond to the reel number you will find if you go to a public library to look at the microfilm. Perhaps all these mysteries are properly resolved if you use FTM in association with ancestry.com.

Be that as it may, I have to admit that in this case, ancestry.com is doing a pretty good job of describing where the data is actually coming from for this particular database. But my experience for many of their databases is that it's extremely difficult to tell from their source citation information where their data actually came from.

Jerry

#30 Jerry Bryan

Jerry Bryan

    Advanced Member

  • Members
  • PipPipPip
  • 3589 posts

Posted 04 June 2013 - 11:44 AM

I continue to make progress with my Extreme Splitting project. I'm continuing with the main outline of the plan and I'm also making a few minor tweaks to improve the plan. One of the minor tweaks is a making decision to try to have at least one media file associated with every single Master Source no matter what. Such an absolutist decision is very left-brained and is perhaps overly obsessive/compulsive and is perhaps not sufficiently creative and flexible. But if I can make this particular tweak to the plan work, then I think it would help to enforce a certain discipline on myself about always recording sources properly.

My definition of a "media file" is actually pretty broad. In addition to JPG type images, it could include MP3 or MP4 recordings of personal interviews, it could include DOCX or RTF or PDF documents that people have sent me, it could include GPX files containing GPS waypoints for points of interest (family homesteads, cemeteries, and even individual grave markers), it could include EML files containing e-mails that people have sent me, etc. But sometimes it's hard to know exactly what an appropriate media file might be.

For example, what if I find an obituary as text at an online site rather than as newspaper microfilm at a library? For newspaper microfilm, the media file that I need is clearly a photocopy from the microfilm. And even if what I find online is images from newspaper microfilm, then the media file for the obituary is clearly some kind of download of the image from the microfilm of the original newspaper. But when what I find online is text, what can I use as the media file?

For obituaries stored online as text, what I have been doing is copying and pasting the text from the online source to the Source Text area of the Master Text area of the Master Source. I have also been trying to construct an appropriate footnote sentence for the citation that makes me happy and hopefully would also make the citation police happy. I've been happy with the results, but there is no media file. So what am I to do about my new rule for myself that every Master Source in RM shall have a media file?

The following (and rather obvious) solution finally occurred to me. I can just make a screenshot of the obituary as I see it online. It's not possible to extract the text of the obituary from the screenshot without OCR. But I don't need to OCR the screenshot because I can already just copy and paste the text. And saving the screenshot provides excellent documentation of where I got the data, including the full URL of where I found the data. Problem solved! Happy camper! Etc.!

However, I now have several dozen of my new style Master Sources - coming up on maybe a hundred of them or so. Before the problem gets any bigger, I want to go back and make sure that every single one of my existing new style Master Sources is associated with at least one media file. So how do I find out which of the new style Master Sources are and which are not associated with at least one media file? The obvious solution would be to run a report. But RM has no reports whatsoever that I can find that even remotely address this issue. The only solution I found within RM itself is to edit every single Master Source manually and to click on the Media tab manually. That's an extremely labor intensive, painful, and error prone process. The absence of any reasonable way to deal with this issue within RM helps to exemplify why I think media management and source management are such weak areas of RM.

I have been trying not to write SQLite reports and to do everything within RM itself. This is a case where there was no choice, and I spent about ten minutes to write a simple little SQLite report that told me exactly what I wanted to know. Here's a link to a part of the output from the report: http://www.screencast.com/t/sj1bdNRtK

In the report it's extremely easy to see which Master Sources have a media file and which do not. For example, the obit for Mary Elizabeth Rader Cox is the first item that is missing a media file. RM's source and media management tools very much need this sort of functionality.

Jerry

#31 TomH

TomH

    Advanced Member

  • Members
  • PipPipPip
  • 6254 posts

Posted 18 June 2013 - 02:43 PM

Further to this most interesting discussion, I have started a page of supportive SQLite queries at Sources - Adventures in Extreme Splitting.

Tom user of RM7550 FTM2017 Ancestry.ca FamilySearch.org FindMyPast.com
SQLite_Tools_For_Roots_Magic_in_PR_Celti wiki, exploiting the database in special ways >>> RMtrix-tiny.png app, a bundle of RootsMagic utilities.


#32 TomH

TomH

    Advanced Member

  • Members
  • PipPipPip
  • 6254 posts

Posted 19 June 2013 - 10:08 PM

The first query I put up converts Source Templates to support extreme splitting. Today, I have added a script that converts citations and sources to extremely split duplicates of the originals, along with the source templates they use. It's a brutish converter that needs a lot of refining but don't let that give extreme splitting a bad rap! Maybe worth a try to see how footnotes come out and whether there are real advantages.

Tom user of RM7550 FTM2017 Ancestry.ca FamilySearch.org FindMyPast.com
SQLite_Tools_For_Roots_Magic_in_PR_Celti wiki, exploiting the database in special ways >>> RMtrix-tiny.png app, a bundle of RootsMagic utilities.


#33 TomH

TomH

    Advanced Member

  • Members
  • PipPipPip
  • 6254 posts

Posted 25 June 2013 - 10:14 PM

This post could easily go under the FamilySearch Family Tree discussion equally well as being a continuation of Jerry's argument in favour of extreme splitting.

The RootsMagic - FamilySearch Family Tree interface seems to be another reason for extreme splitting. Take any source created in RootsMagic, templated or Free Form and transfer it to FSFT. Then import the person and the source to a RM database. What you get back is a Free Form Source with the Source Details (Page number) field blank and the Master Source fields Footnote, Short Footnote and Bibliography containing the original full footnote sentence. The title of the source is the same as the original. You get back the original Master Source Text, the first of the original Master Source webtags and nothing else! Neither Comments values, no Research Notes, no Detail Text Webtags, no Media, no Quality, no Repository. Lumped sources in RootsMagic become extremely split en route to FSFT and stay that way on the way back, except that key information is lost. Extremely split sources are already nearly optimal for transfer, except for the loss of Source Comments, secondarily entered WebTags, ...

Here's an example of what came back from the transfer of two sources for Annie Florence Holden-[KCH9-5HR] from RootsMagic to FSFT and back. The original source template was a copy of "Website (with multiple databases)".

Returned from the original lumped source:
Ontario Marriage Registrations - Rootsweb - The Ontario Vital Statistics Project (lumped)

Citation: Mary Crandall, "Ontario Marriage Registrations," database, Ontario Vital Statistics Project, The Ontario Vital Statistics Project (http://homepages.roo...c/thisisit.htm: accessed 21 November 2010), Part 2: 1887 Ontario Co. 8902-87 John Edgar ALLEN & Annie Florence Holden; citing Part1: 1800 - 1925 Marriages arranged by county, from sources other than post-1869 civil registrations; Part 2: Ontario Civil Registration from 1869 to 1927, from Ontario Archives, microfilm MS 932.

Click to view record (this was added manually to the FSFT source as there was originally no webtag nor would one at the Detail Text level have made it to FSFT)

Returned from the original extremely split source batch generated by http://sqlitetoolsfo...s and Citations:

*Ontario Marriage Registrations - Rootsweb - The Ontario Vital Statistics Project (extremely split)

Citation: Mary Crandall, "Ontario Marriage Registrations," database, Ontario Vital Statistics Project, The Ontario Vital Statistics Project (http://homepages.roo...c/thisisit.htm: accessed 21 November 2010), Part 2: 1887 Ontario Co. 8902-87 John Edgar ALLEN & Annie Florence Holden; citing Part1: 1800 - 1925 Marriages arranged by county, from sources other than post-1869 civil registrations; Part 2: Ontario Civil Registration from 1869 to 1927, from Ontario Archives, microfilm MS 932.

Source text: || 8902-87 -: John Edgar ALLEN, 27, farmer, Brock twp., Whitby twp., s/o John & Mary, married Annie Florence HOLDEN, 21, Prince Albert, Whitby, d/o James & Orilla, witn: Amelia ALLEN & Frank ELLIKER, both of Whitby twp., 3 Feb 1887 at Whitby (Note that the text after the "||" was moved from Detail Text to Source Text by the converter so it made the round trip)

Click to view record (this was added manually to the FSFT source as there was originally no webtag but had there been one for the original Master Source, it would have returned here)

Tom user of RM7550 FTM2017 Ancestry.ca FamilySearch.org FindMyPast.com
SQLite_Tools_For_Roots_Magic_in_PR_Celti wiki, exploiting the database in special ways >>> RMtrix-tiny.png app, a bundle of RootsMagic utilities.


#34 Jerry Bryan

Jerry Bryan

    Advanced Member

  • Members
  • PipPipPip
  • 3589 posts

Posted 26 June 2013 - 07:55 AM

I started out with Extreme Splitting of sources as sort of an experiment to try to solve some difficult problems in managing sources in RM. I was so pleased with the results that I have become extremely committed to the idea. The only two flies in the ointment have been short footnotes and bibliographic entries, neither of which works with Extreme Splitting. Actually, I'm not much worried about short footnotes. Short footnotes are only used when a report is created with what I will call "true footnotes" - those that appear at the bottom of each page. I think that true footnotes are extremely dysfunctional. Endnotes - that is, "footnotes" that all appear together at the end of a report - work so much better than true footnotes that I will surely never use true footnotes. I still think I would like to be able to use bibliographies on occasion, and Extreme Splitting is pretty unfriendly to bibliographies.

But I've come to realize that there is another fly in the ointment of Extreme Splitting, namely Source Quality. A given source citation can be used for more than one purpose, and the quality of the source citation can vary depending on its purpose. For example, a death certificate usually may be cited as evidence both for a death fact and for a birth fact. For a death fact, a death certificate is usually primary. For a birth fact, a death certificate is usually secondary. RM supports this idea by having Source Quality associated with the Source Details part of a source citation rather than with the Master Source part of a source citation. Hence, RM's Source Quality feature cannot be used in association with Extreme Splitting.

I've never used the RM's Source Quality feature for sources and their citations, but I've recently been considering doing so. I actually think the value of using this feature is actually pretty limited. The Source Quality data doesn't appear in narrative reports, so it's mostly for my own personal use. And I usually know what the quality issues are with a source and its citations. If there are any quality issues that I consider important enough to bring forward in narrative reports, I include any necessary commentary in a fact note that does appear in narrative reports. So my decision is basically that Extremely Splitting is so important to my future strategy in documenting my genealogy that I'm willing to abandon all hope of using RM's Source Quality feature.

Jerry

#35 TomH

TomH

    Advanced Member

  • Members
  • PipPipPip
  • 6254 posts

Posted 26 June 2013 - 08:45 AM

Were RootsMagic to incorporate a mechanism to include Research Notes et al in what is uploaded to FSFT, then the particular advantage I demonstrated above to employing Extreme Splitting in the RM database would be nullified and one could continue to use lumped sources with their particular strengths and weaknesses. If RM never does so, then maybe there will be some merit in converting to extremely split sources and some more complete version of my conversion script at http://sqlitetoolsfo...litting#Convert Existing Sources and Citations would make it easy.

I think you give Source Quality short shrift in the context of extreme splitting. Even with extreme splitting there is a Source Quality to each citation so when you copy/paste a source from one fact to another, you can set the two Source Quality's independently.

Tom user of RM7550 FTM2017 Ancestry.ca FamilySearch.org FindMyPast.com
SQLite_Tools_For_Roots_Magic_in_PR_Celti wiki, exploiting the database in special ways >>> RMtrix-tiny.png app, a bundle of RootsMagic utilities.


#36 APerson

APerson

    Advanced Member

  • Members
  • PipPipPip
  • 267 posts

Posted 27 June 2013 - 12:01 PM

It's really a shame that Tom continues to promote his faulty "study" in which he promotes his free style template while grossly misrepresenting "Simple Citations." A full description of his rather bizarre "study" and his equally bizarre conclusions may be found here: http://www.simplecit...Comparison.html

#37 TomH

TomH

    Advanced Member

  • Members
  • PipPipPip
  • 6254 posts

Posted 27 June 2013 - 09:42 PM

Jeff,
1. How does your post flow from anything I have said in this discussion?
2. How does it even relate to the topic of the discussion?

Non-sequitur and off-topic characterize your post. If you wish to resume a discussion about source templates, continue an appropriate thread or start one rather than cluttering up this very different and informative discussion.

Jerry's exploration of extreme splitting of sources has revealed a number of (surprising) advantages over moderate lumping which is inherent in the design of most built-in source templates, including Free Form, plus custom ones such as yours and mine. Any source template with fields on the input form in the green, Source Details area is one that promotes or invites lumping. The higher the proportion of fields in green, the more the lumping (fewer Master Sources); the converse is true for splitting.

The advantages are in RootsMagic's behaviour and its interface to external systems, the latest of which is the transfer of sources to FamilySearch Family Tree which fails to carry over Research Notes and the Citation WebTag. Only the Master Source Text and WebTag make the leap. Thus lumping to any degree is guaranteed to fail; only the ultimately split sources can transfer correctly within the constraints of FSFT. This seems to be yet another example of incomplete fulfillment of a design feature, think Place Details, Shared Events, Source Templates...

To better examine the effects of extreme splitting with one's own database, I have been putting effort into a SQLite script to convert lumped sources into ultimately split sources. Concomitant with that is the conversion of the corresponding source templates into custom templates with all data fields moved up into the yellow, Master Source area. You will find that the hyperlink I provided in my previous post goes to a page that describes and makes available these conversion scripts and says nothing about my unrelated study of source templates.

Such conversion might be one time and extreme splitting continued going forward or it might be used per occasion for export via GEDCOM to 3rd party systems or for transfer to FSFT. It may also help solidify user opinion about what needs to be changed in the RootsMagic source transfer process that could obviate having to do the conversion at all.

The conversion script should work equally with your "Simplified Citations" as it does for all other source templates.

Tom user of RM7550 FTM2017 Ancestry.ca FamilySearch.org FindMyPast.com
SQLite_Tools_For_Roots_Magic_in_PR_Celti wiki, exploiting the database in special ways >>> RMtrix-tiny.png app, a bundle of RootsMagic utilities.


#38 APerson

APerson

    Advanced Member

  • Members
  • PipPipPip
  • 267 posts

Posted 29 June 2013 - 12:07 PM

. . .
That said, my test was with the Census, US Fed - Online images template and an edited copy of it, as I used in Source Templates, A Comparative Example.pdf. In practice, I cannot imagine splitting to the level of the Person of interest | Line number, nor even to the page number. . .
.


Simple, as you continue to refer to your bogus "study" that misrepresents Simple Citations in this thread (see above), I have pointed out the significant problems with your bizarre "study" here:

http://www.simplecit...orm_Review.html

and

http://www.simplecit...Comparison.html

As I clearly noted on first of the above links, your free-form template is an example of splitting, not lumping.

#39 Jerry Bryan

Jerry Bryan

    Advanced Member

  • Members
  • PipPipPip
  • 3589 posts

Posted 03 July 2013 - 06:39 PM

I made a screencast of using the Extreme Source Splitting technique with a grave marker photograph as a source. I could have used any number of other types of sources as an example, but this one was convenient. Unfortunately, it's about 9 minutes long but it moves along pretty well.

The URL is http://screencast.com/t/R1Q28kkYlhAa

Jerry

#40 TomH

TomH

    Advanced Member

  • Members
  • PipPipPip
  • 6254 posts

Posted 03 July 2013 - 08:36 PM

That's a fine demo of how you create a complete citation of a new source, tagged with media, completely within the Source Manager without ever drilling down to Source Details, Detail Text, etc.

I like your naming convention: sourcetype: personname; supplementarydescription. I have been working on a script that converts citations of lumped sources to what I might call "ultimately split sources" (possibly going beyond "extremely split"!) and am almost finished but for naming the new sources. I have Person sources now being named oldmastersourcename: personname (Person) with the intention of extending that pattern to oldmastersourcename: personname (eventname). I'm not sure how well that will work from a user perspective but it is programmatically appealing. Any comment?

The goal for this script is to make it easy to transform an existing database to one with all sources ultimately split so that anyone (with SQLIte basics) can see how this affects outputs from their database. I'm expecting to see GEDCOM source output that matches report endnotes, regardless of which source template is used.

Tom user of RM7550 FTM2017 Ancestry.ca FamilySearch.org FindMyPast.com
SQLite_Tools_For_Roots_Magic_in_PR_Celti wiki, exploiting the database in special ways >>> RMtrix-tiny.png app, a bundle of RootsMagic utilities.