Source Templates Formatting Lost

RootsMagic
1

When I started using RM10 earlier this year, I was disappointed to see the gibberish that became of my sources when I treeshared to Ancestry, so I stopped. Over the last couple of days, I played with RM’s gedcom exports. I was extremely disappointed to see the exact same behavior importing the gedcom into both Ancestry and MacFamilyTree. My sources are useless.

Is this something I’ve done wrong with all of my templates? Is it because I lump my sources?
Here’s an example of a census.

Master source for the 1880 census.

master
master1324×1096 152 KB

Citations for the 1880 census

source
source1920×947 153 KB

2

@tinksquared Sorry I am NOT understanding why you say this is gibberish–I think it looks good–some might say something abt Evident Explained standards which I am NOT a fan of BUT it looks good in my opinion..
Only problem is that you copied the wrong URL-- I typed out the URL above and it did NOT work ( But I could have made a mistake) --I think you copied the search URL-- you should have either copied the information page ( 1st page)
https://www.ancestry.com/search/collections/6742/records/19226707?tid=&pid=&queryId=a691c053-ccb6-413c-9a3f-2ae2a603fe87&_phsrc=Sok5&_phstart=successSource

or the ACTUAL IMAGE URL-- which Ancestry is NOT letting me copy for some reason BUT other companies do allow it..

I personal prefer the information page as sometimes I want to look at something else besides what I pulled up..

One other thing in the 1880 census, there are 4 parts to every page number so this one should say 447 C
edit–when I copy a URL that I want to save, I always open a new window/ tab and paste the copy in to see that it works right–sometimes they do NOT…

3

Yes it looks fine in RM. But when I export it anywhere; gedcom, ancestry, etc. It’s complete gibberish.

4

I don’t know what you mean by gibberish so I can’t speak to that. However the URL can actually be shortened in this case. Everything after the ? can be eliminated and it will still take you to where you want to be. So use ‘https://www.ancestry.com/search/collections/6742/records/19226707’ instead.

One thing to note is that the various websites and program all have their own data models so everything lines up all the time.

5

The citation looks like this in Ancestry (treeshare) or gedcom export.

Title

1880 United States Federal Census, , pop.sch., , ; , [database with images], : accessed ; citing (NARA microfilm publication T9, 1,454 rolls). Records of the Bureau of the Census, Record Group 29. National Archives, Washington, D.C..

**
Transcript**

Name: Lawrence BurgmeierAge: 47Birth Date: Abt 1833Birthplace: BadenHome in 1880: Spencer, Jennings, Indiana, USADwelling Number: 217Race: WhiteGender: MaleRelation to Head of House: Self (Head)Marital Status: MarriedSpouse’s Name: Dena BurgmeierFather’s Birthplace: BadenMother’s Birthplace: BadenOccupation: FarmerName: Dena BurgmeierAge: 45Birth Date: Abt 1835Birthplace: BavariaHome in 1880: Spencer, Jennings, Indiana, USADwelling Number: 217Race: WhiteGender: FemaleRelation to Head of House: WifeMarital Status: MarriedSpouse’s Name: Lawrence BurgmeierFather’s Birthplace: BavariaMother’s Birthplace: BavariaOccupation: HousekeeperName: John BurgmeierAgeL 22Birth Date: Abt 1858Birthplace: IllinoisHome in 1880: Spencer, Jennings, Indiana, USADwelling Number: 217Race: WhiteGender: MaleRelation to Head of House: SonMarital Status: SingleFather’s Name: Lawrence BurgmeierFather’s Birthplace: BadenMother’s Name: Dena BurgmeierMother’s Birthplace: BavariaOccupation: Working on FarmName: Lawrence Jr. BurgmeierAge: 20Birth Date: Abt 1860Birthplace: IllinoisHome in 1880: Spencer, Jennings, Indiana, USADwelling Number: 217Race: WhiteGender: MaleRelation to Head of House: SonMarital Status: SingleFather’s Name: Lawrence BurgmeierFather’s Birthplace: BadenMother’s Name: Dena BurgmeierMother’s Birthplace: BavariaOccupation: Working on FarmName: Bertha BurgmeierAge: 14Birth Date: Abt 1866Birthplace: IllinoisHome in 1880: Spencer, Jennings, Indiana, USADwelling Number: 217Race: WhiteGender: FemaleRelation to Head of House: DaughterMarital Status: SingleFather’s Name: Lawrence BurgmeierFather’s Birthplace: BadenMother’s Name: Dena BurgmeierMother’s Birthplace: BavariaOccupation: HomeworkerName: Henry BurgmeierAge: 10Birth Date: Abt 1870Birthplace: IndianaHome in 1880: Spencer, Jennings, Indiana, USADwelling Number: 217Race: WhiteGender: MaleRelation to Head of House: SonMarital Status: SingleFather’s Name: Lawrence BurgmeierFather’s Birthplace: BadenMother’s Name: Dena BurgmeierMother’s Birthplace: BavariaOccupation: StudentName: Jacob BurgmeierAge: 7Birth Date: Abt 1873Birthplace: IndianaHome in 1880: Spencer, Jennings, Indiana, USADwelling Number: 217Race: WhiteGender: MaleRelation to Head of House: SonMarital Status: SingleFather’s Name: Lawrence BurgmeierFather’s Birthplace: BadenMother’s Name: Dena BurgmeierMother’s Birthplace: BavariaOccupation: StudentName: Mary BurgmeierAge: 5Birth Date: Abt 1875Birthplace: IndianaHome in 1880: Spencer, Jennings, Indiana, USADwelling Number: 217Race: WhiteGender: FemaleRelation to Head of House: DaughterMarital Status: SingleFather’s Name: Lawrence BurgmeierFather’s Birthplace: BadenMother’s Name: Dena BurgmeierMother’s Birthplace: BavariaCannot Read: YCannot Write: Y

6

Those joined pieces of text are definitely related to where a carriage return previously existed (ie. attempted formatting of notes or missing sentence template spacing between parameter elements), maybe?

7

I don’t know. I wish I did. But the way RM templates export is awful. Unless I’m doing something wrong, which is why I posted an example.

8

as @kfunk explains – the stuff after ? is often used for tracking etc or has parameters for specific info. For Citation only the “root” Url is likely needed – you might want the longer URL in some cases for webtag or such to get directly to the desired item

9

Thanks. But completely irrelevant to my question. I’m not worried about the URL’s, although that is great to know and much appreciated.

10

Agreed much garbage comes form Ancestry they way things get shared/mapped.
This was true with FTM also. So after I moved to RM I do not shared anything form ANC to RM.

I do it manually (anything to do with media, source and the like). Do most things manually.
Family Search is much better over all with transfer but that depending on how clean the entry was — meaning it will match what is on FS (good or bad as for as location & so on)

11

Sorry to say BUT in my opinion, it has nothing to do with the way RM templates export but rather what the receiving program does with the info- I tried opening a gedcom from RM with another genealogy program and Ancestry— both pretty much mangled ( or gibberish if you prefer) most of my sources --a couple turned out fairly descent BUT NOT any where close to what it looks like in RM–Ancestry definitely has it’s own way of doing sources…

Unfortunately this has been a problem for a very long time-- when I first switched from one genealogy program to another by way of gedcom, it took all my sources and turned them into free form sources and basically stripped out all the info I had added.

Have you tried adding a source to MacFamilyTree and seeing how it transfers as a gedcom–think you will see the same results BUT only thing I can think of for now…

2 Likes
12

I cannot speak to gedcom imports to ancestry but I share your frustration with how RM text appears via treeshare. My sense is that this is much more an Ancestry issue than a RM one and it’s a matter of personal choice as to how much you want to configure your RM data to optimize the way it will appear after a treeshare.

It’s been a few years since I last verified the specifics of how field data transfers but as far as I know the following still applies:

  • You can get a proper Title field when treesharing from the “Ancestry Record” source template. The footnote field of the freeform template also maps to the Title field on ancestry. All other templates concatenate the template fields with comma delimiters as your example shows. Since RM maps 2 source templates to the title field, the concatenation of fields is clearly something they could change. You could consider a feature request to address the Title field mapping (perhaps if a Title field appears in the RM source Template then map 1:1, else concatenate. Some consideration/discussion would need to occur regarding the treeshare behavior of sources from the freeform template.)

  • The formatting characters in multiline text get stripped out. My understanding is that this occurs on the Ancestry side of the API. (It used to be worse, they used to not display any characters after a “<” but that’s no longer the case.)

  • If the citation was originally treeshared from Ancestry to RM (ie is in an “Ancestry Record” source template and has an associated AncestryRecordTable anID value), then a subsequent treeshare of that source, or reused sources derived from that source, will look like an original Ancestry Source. In other words the source record will look the same as it did before the initial treeshare to RM. So, it will look nice but your RM edits won’t get displayed. In this case the only way to have your comments/edits appear in ancestry is to add them to a fact note in RM. The fact note will treeshare up as a unique element tied to the event and display in the “Other Sources” part of the ancestry webpage and, again, all new line characters, double spacing, etc. gets stripped out.

1 Like
13

I’m pretty sure I know the answer. It isn’t because you are a lumper, per se, although that sort of has something to do with it. When you export citations out of RM that are generated by templates, RM has to convert the citations to what I will call “GEDCOM standard” for lack of a better term. In doing so, all the Master Source variables are grouped together, and all the Source Details Variables are grouped together. And when I say RM “has to”, it really does have to. The receiving system may not understand RM’s source templates. The only thing available to the receiving system is going to be “source variables” as one item and “citation variables” as one item. These may or may not be be mangled as compared to the way they appeared in RM. It all depends on how the citation sentence is set up.

This is a well known problem, and there really isn’t much RM can do about it. Well, the source templates could have been designed so that all the “master source” variables were to the left of all the “source details” variables in the citation sentence. But then they probably wouldn’t have been compliant with Evidence Explained. The template you are using in your example is pretty egregious in not having “master source” variables to the left of all “source details” variables.

I’m an extreme source splitter and I use custom source templates of my own design. As a result, my citations are never mangled, no matter the system to which they are exported. And I’m sure that I am not 100% Evidence Explained compliant. Here are a couple of examples. All variables in the source (AKA master source).

U.S. Federal Census: 1940, Anderson County, Tennessee, Dist. 9, Enumeration District 1-14, page 243a, ancestry.com (1940 U.S. Census) viewed on 24 November 2014.

Notice that I don’t include family ID or dwelling ID or name of the person or anything like that. I cite to the census page. Period. There are nearly always multiple people per family and often multiple families on the same page that have the exact same citation. That’s what make sense to me. I don’t use repositories per se because there are usually so many places to find the same image. But I do make note of when and where I found the image. But the Web site is not entered as a repository.

Death Certificate: Peters, Alva Edward; Certificate 27359; Edgemoore, Civil District 9, Anderson County, Tennessee, 1942; ancestry.com (Tennessee, Death Records, 1908-1958); viewed 29 September 2015.

To me, the important thing is that this is the death certificate for Alva Edward Peters, so that’s what I put first in the sentence. The rest is fluff as to where to find it (well, where to find it actually is important, and is not just fluff). But a typical source template controlled citation is going to put the person’s name in the citation (AKA Source Details) rather than in the source (AKA Master Source). So my sentence which is logical to me only works for export if I’m an extreme source splitter. It wouldn’t work for export if the deceased person’s name was in the citation (AKA Source Details).

There is one more minor issue that is related. The sentence for a source template often has fixed text, not controlled by template variables. All such fixed text needs to be to the left of all citation variables (AKA Soure Details variables) to avoid the same kind of mangling.

My discussion is moot of you are exporting to RM or if you are exporting to software that does understand RM’s source templates.

3 Likes
14

MacFamilyTree does not mangle my sources when exporting to gedcom.

15

Great explanation thank you. It is definitely a RM Template issue.

16

That can be accurate – I think it many cases – the way the receiving program “Maps” (or remaps ) the diff fields can definitely alter the results

17

Could be, but again, I can export gedcom sources made with templates in MacFamilyTree to Ancestry and they look just fine. MFT does export in gedcom 7. I don’t know if that makes a difference, as RM cannot.

18

I feel like I’m repeating myself a little bit, but let’s suppose that we are exporting sources from RM to GEDCOM without the option Extra details (RM specific). This is the option you might use if the software that was going to import the GEDCOM could not understand RM’s source templates. The relevant GEDCOM tags are as follows.

image
image977×303 45.9 KB

In particular, the only two places the data from RM’s source templates can go are the 1 TITL tag which can contain all the master source variables and all the fixed text, and the 2 PAGE tag which can contain all the source detail variables. So without the Extra details (RM specific) option for exporting GEDCOM, this is what I mean by the fact that RM has no choice but to mangle the citation sentence unless all the master source variables and all the fixed text in the citation sentence in the source template are to the left of all the source details variables. This has nothing to do with the software that is reading the GEDCOM. The problem at the stage of the game is 100% inside of RM.

I would further repeat that some of RM’s source templates do have all the master source variables and all the fixed text in the citation sentence in the source template to the left of all the source details variables. Such source templates will never cause citation sentences they control to be mangled on GEDCOM export. But some of RM’s source templates do not have this characteristic. In other words, some of RM’s source templates have at least one source details variable that is to the left of at least one master source variable or fixed text or both. Such source templates will cause the citations sentences they control to be mangled on GEDCOM export when the Extra details (RM specific) option is not used.

Again, I repeat: this has nothing to do with the receiving system. This issue is 100% within RM. So, aha! Maybe we could solve the problem by always exporting GEDCOM while using the the Extra details (RM specific) option. But not so fast. Because now it does depend on whether or not the receiving system can process those Extra details (RM specific) options. Namely, can the receiving system process RM’s source templates?

If you do export GEDCOM using RM’s Extra details (RM specific) option, what actually happens is that it puts all your sources and citations into the GEDCOM twice. One copy of your sources and citations goes into the GEDCOM as already described where citation sentences can be mangled for some source templates and not mangled for other source templates. The other copy of your sources and citations goes into the GEDCOM with full information about the source templates. In this case, it totally does depend on what the receiving system does. If it does not not understand RM’s source templates, it will use the GEDCOM just as if Extra details (RM specific) option had not been specified. In other words, some of the citation sentences may be mangled even though you used the Extra details (RM specific) option in RM. But if the receiving system does understand RM’s source templates, your citation sentences will not be mangled by the receiving system.

But there are other kinds of export from RM other than GEDCOM. For example, there is export to Ancestry via TreeShare and there is export to FamilySearch via FamilySearch Central. What happens in this case? Well, we can’t see the data going to Ancestry and FamilySearch the same way we can look at a GEDCOM file. But it is clear that neither Ancestry nor FamilySearch can understand RM’s Source Templates. In that case, RM has no choice but to organize the data the same way as it does when it exports GEDCOM without the Extra details (RM specific) option. In other words, the citation sentences for some source templates will be mangled.

So what is an RM user to do?

  1. Maybe you don’t care because you are not exporting anywhere that results in mangled citation sentences.
  2. You can use only RM’s free form source template.
  3. You can use some alternatives to RM’s source templates that are floating around that might not be susceptible to having their citation sentences mangled upon export.
  4. You can use only those RM source templates that are not susceptible to having their citation sentences mangled upon export. This can be hard to do.
  5. You can develop your own source templates that are not susceptible to having their citation sentences mangled upon export. You could develop them from scratch, or you could develop them by copying some of RM’s source templates and modifying them.
  6. You could become an extreme source splitter.

I’m probably leaving out a possibility or two, but I think that is the main thrust of it.

1 Like
19

I’ve made the decision to go back to MacFamilyTree as my master DB. I will continue syncing a gedcom from MFT to RM to use some of the features.