Question How to correctly denote a website, Facebook profile or URL as a source?

After migration to 2.1.7 webtrees tells me that I have a huge bunch of errors like these:

How should I correctly denote that given source is entirely and exclusively based on some URL?

I have browsed SOURCE_CITATION and SOUR_RECORD pages in the GEDCOM-5.5.2, but found no clue how to do this correctly.

An example:

Unfortunately this is not an option for me at all. I have sometimes 50+ facts based on single source. I cannot imagine pasting the same URL fitty times, because this creates an overwhelming data redundancy and makes updating such tree a horrible experience.

I find quite hard to believe that even in newest version of GEDCOM there is no support for entirely URL-based sources and that what you propose is the only way of solving this problem.

Anyway, I need find a way to store an URL or URLs one time per each source and to not repeat them each time per each fact.

In the worst case scenario, if anyone else cannot come out with any better solution, I will use NOTE_STRUCTURE or TEXT_FROM_SOURCE element from SOUR_RECORD tag or from SOURCE_CITATION .

This seems as weird as it can be, but I don't see any other options. And I certainly don't see the possibility of repeated URL for the same source web page per each event, that source web page describes.

Unfortuntely, GEDCOM does not allow us to create a "citation object", which we could re-use.

What is your definition of "Source"?

In GEDCOM we have three levels of sourcing data.

Here is one example of how I would use GEDCOM to source data from a website.

1. The Repository - The Place that the data is found, this equates to a Library. (This could be a repository like Ancestry.com)
2. The Source - The Document that the data is contained in, the equates to a Book. (This could be a census, by year and place)
3. The Citation - The location that the data itself was read from, this equates to the Line/Paragraph/Section in the source book. (This could be a URL page with and image of the census page)

If 50 facts came from one paragraph or "Page" of a website then yes you would need to duplicate the page reference in 50 citations, as stated above citations are not normalized into one record in GEDCOM.

Trejder said:

I will use NOTE_STRUCTURE or TEXT_FROM_SOURCE element from SOUR_RECORD tag or from SOURCE_CITATION .

You would be out of luck using the "Source_Citation" for either a Note_Structure or Text_From_Source since they exist in the same structure as "PAGE" in my example above!

trejder wrote:

I find quite hard to believe that even in newest version of GEDCOM there is no support for entirely URL-based sources and that what you propose is the only way of solving this problem.

Anyway, I need find a way to store an URL or URLs one time per each source and to not repeat them each time per each fact.

I think it needs same differentiation whether your URL belongs to the source or whether it is part of the source citatation in a fact or an event:

If the URL belongs to the source, we can regret that GEDCOM does not offer SOUR:WWW. However, there are still possiblilites to deal with the issue. For example, I have URLs to the online church records of data.matricula-online.eu/ . In this case, I put the URL into the call number, because webtrees nicely handles URLs in call numbers; you can directly click the link!

If you don't like this approach, you alternatively might want to put the URL into a note.


If your URL shall be part of the source citation, it is different. The idea of the citation is that it defines the location (e.g. page) where something is found within a source. If your URL points to a certain location within a source, it makes sense to use it in the source citation. For Matricula, I use the URL to the specific page within the church record (in this example: "/?pg=35"):

Like Greg mentioned in his post, we can still regret that GEDCOM does not offer data structures to reuse this kind of source citations. If you might have 50 citations from a certain page in a census record, it would need duplicating the above structure.

Finally, if you do not have a specific URL to point to the location within your source, you do not need the URL at all in the source citation. You can simple put the page number (in this example "Page 35") into the source citation. If the URL points to the source (and not to a location within the source) is not needed in the source citation.

Jefferson49 said:

I think it needs same differentiation whether your URL belongs to the source or whether it is part of the source citatation in a fact or an event:

If the URL belongs to the source, we can regret that GEDCOM does not offer SOUR:WWW. However, there are still possiblilites to deal with the issue. For example, I have URLs to the online church records of data.matricula-online.eu/. In this case, I put the URL into the call number, because webtrees nicely handles URLs in call numbers; you can directly click the link!

Yes, this is why I asked, "What is you definition of a 'Source'?"

Depending on how you define "source" determines how to use GEDCOM. Most people wrongly say "source" when they mean "citation" as in where in a book (a source) they found the data and they "cite" the location (a page number) in that book. So I may have misunderstood what you were asking about.

- A "Citation" is always associated with a specific Fact/Event where a source is a stand-alone object. Normally this would be a "Page" on a website.

- A true "Source" is a stand alone object that can be cited by multiple Fact/Events and can be found at a Repository (aka a web site). This source is normally something like a census, church document, manuscript, or other collection of information.

As Jefferson49 indicated the URL for the "Source" is the REPO:CALN tag of the Source_Record and generally points at the home page of the website or some starting point of the set of census images on the web site.

The hitch come in where the website is one page and it goes on and on with no link-able headings that can be used to "Cite" a Fact/Event.

I like to maintain all of my citations (the location of the data) in the Source_Citation structure, this way all reports and displays need to look only in one place to display the location information.

Thank you both for an extensive and detailed information and answer.

@Jefferson49 Each of your proposals is more or less similar to @norwegian_sardines initial proposal and assumes that an URL (or other element in your other examples) is put directly in the FACT or EVEN, not in the SOUR. That's not an option for me, for the reasons already expressed.

Please, note that I don't judge here that your idea is wrong or, if my request is possible. As Greg has already explained, it might be impossible to store an URL directly on SOUR element though from my perspective this is as intuitive as it can be. I am not judging that your approach is wrong, I am only responding that it does not fit my needs.

@norwegian_sardines The answer (and an example) is as simple as it can be. I am using:

• Facebook (literally) as a REPO
• Given Facebook profile as a SOUR

I have many facts (birth date, marriage date, family members, children, multimedia objects (mainly photos) all sourced by that single Facebook profile.

And I want to have an URL attached to that SOUR, because it is a single URL for single source.

As I said, unless you come with something better that still fits my needs, I will probably end up with having URL placed as a NOTE or TEXT on SOUR.

Maybe use the PUBL subtag.

-Facebook (literally) as a REPO
-Given Facebook profile as a SOUR

Yes, this is how I would layout this relationship.

Repository_Record would be information about the Facebook site in general, not the page.

Source_Record would contain specific information about the Facebook page that you took the information off. I would use the REPO.CALN tag to document the Facebook page URL.

Source_Citation would be used to document specific entries that cite what you found on the specific page. If you can use the “permalink” for a specific article about the information you took about an individual fact/event. This URL would go in the PAGE tag.

trejder wrote:

@Jefferson49 Each of your proposals is more or less similar to @norwegian_sardines initial proposal and assumes that an URL (or other element in your other examples) is put directly in the FACT or EVEN, not in the SOUR. That's not an option for me, for the reasons already expressed.

Maybe, my explaination was too detailed and showed too many additonaly possibilities, which can be misleading. I understood now that you mainly look for a GEDCOM structure where to put your URL within a source.

Like norwegian_sardines, I think that a good place is REPO:CALN. The call number (CALN) says where you find a source in a repository. This is pretty close to what your URLs do, i.e. a Facebook URL says where you will find the source (the Facebook profile) within the repository (Facebook).

webtrees has already recognized that URLs are a typical use case in call numbers. Therefore, URLs in call numbers are shown as links, which can be directly clicked.

It seems that using SOUR.REPO.CALN is the proposal, I was looking for.

Thank you, both, for your time spent on solving this issue!

The answer (and an example) is as simple as it can be. I am using:

Facebook (literally) as a REPO
Given Facebook profile as a SOUR

I have many facts (birth date, marriage date, family members, children, multimedia objects (mainly photos) all sourced by that single Facebook profile.

And I want to have an URL attached to that SOUR, because it is a single URL for single source.

Why don't you enter "Facebook" as source (SOUR) and the Facebook profile as a source citation (SOUR:PAGE). For this you have to create a source once and change f.i. SOUR:WWW to SOUR:PAGE by search&replace. Under REPO:CALN I would not enter an internet address.

mp wrote:

Why don't you enter "Facebook" as source (SOUR) and the Facebook profile as a source citation (SOUR:PAGE). For this you have to create a source once and change f.i. SOUR:WWW to SOUR:PAGE by search&replace. Under REPO:CALN I would not enter an internet address.

Because this way, I will have all my facts denoted (sourced) as: "Source: Facebook". I don't want that. I want it to have as I have it right now: "Source: Facebook's profile for XX".

wrote:

Dealing with GEDCOM tags can be a bit tricky, but don't worry; I've got you covered. To denote a source based solely on a URL, you can use the SOUR tag followed by a description.


 

Edited!

“Followed by a Description”? I’m not sure what you mean! A Source_Record does not have a tag for “description”.

The common tags are: Author, Title, Publisher, Text from Source, Repository Citation, Source Abbreviation.

If you are talking about an “inline” Source_Citation, then this structure does not use a link to the Source_Record and any additional information must use the “description” tag to store all information about the source and repository of the fact cited! This is an old structure and not very useful (because of the ill defined data structure) in many potential reports!  It also does not make the URL linkable like using the PAGE tag!