Citation templates are none of GedcomX's business.

[jralls]

Citation templates are about displaying citation information in reference notes and bibliographies: They're the province of applications. GedcomX's mission is about passing information, including citation information, between applications, so our concern should be with the citationFields and we should leave citation templates to projects like Citation Style Language.

[fleep]

Has any work been put in into enumerating CitationField types, or are we kicking that off here? Or are we trying to avoid such enumeration at all?

[jralls]

Looks like the intent is to define what a template would look like and leave the definition to others.

Lots of work has been done in a variety of places, but this isn't one of them. ;-) IMO it would be foolish to redo the previous work.

[thomast73]

In line with what @jralls has expressed, we believe there should be a separation between (1) how fields and their data types are defined, (2) how fields values are exchanged, and (3) how field values are brought together to form a rendered citation.

@fleep asks about the enumeration of field types. This is an area that may not be "the business" of GEDCOM X. I have looked into this and I am not in favor of pursing such a list. See the discussion in the GEDCOM X issue forum. [I'm not sure the discussion there represents all of the current thinking (I did not fully review the thread, and it is a bit old), but I am hoping it would be useful in resuming discussions here.]

I expect that GEDCOM X will be concerned with how field values are exchanged. I believe GEDCOM X needs to specify a mechanism for defining fields (a data definition language of sorts?)—likely via a template mechanism of sorts. Perhaps GEDCOM X will also need to pave the path for transforming field values into rendered citations (a language for mapping fields into a rendering citation?)—something that could become part of a template mechanism, though things like CSL might already be usable in this regard. But I do not think it is "the business" of GEDCOM X to review all of the citation needs of all languages and all cultures and then catalog and canonize said knowledge into a standard. Rather, I see GEDCOM X defining an infrastructure for defining/describing an abstraction of a citation (the fields that are needed, how to exchange them, and perhaps options for rendering them), and then allow the genealogical community to develop and curate a "library" or such definitions adjunct to the GEDCOM X standard.

[jralls]

Rather, I see GEDCOM X defining an infrastructure for defining/describing an abstraction of a citation (the fields that are needed, how to exchange them, and perhaps options for rendering them), and then allow the genealogical community to develop and curate a "library" or such definitions adjunct to the GEDCOM X standard.

+1

[fleep]

Again, I apologize for my lack of prior involvement in the ongoing discussions around all this :smile: I am trying to get up to speed on some of this stuff - specifically the boundaries of where GEDCOM X acts as a vessel for structured data-exchange, and how much it tries to elucidate the nature of the data it contains in a predictable way.

I do like the template format. I would suggest that some basic enumeration of field types would be useful, but perhaps this is not the purview of GEDCOM X. As a person who runs a web application in the genealogy space, I would like to easily be able to consume (for example) references to URLs or API resources where information was shared from or cited or imported.

Maybe I'm being wrongheaded about the usefulness of citations in an application aside from human-consumption? I'm open to the possibility that I'm overthinking it.

[fleep]

Though I suppose if SourceTemplates (or some other authority) defines a few templates generally-accepted by the application-building community, this would address quite a bit of that.

[jralls]

Though I suppose if SourceTemplates (or some other authority) defines a few templates generally-accepted by the application-building community, this would address quite a bit of that.

Do you mean SourceTemplates.org? They're the same company as AncestorSync. They're nice guys, they've been at all of the RootsTechs and the last two NGS Family History conferences, but they're not an "authority" in any sense.

[thomast73]

Though I suppose if SourceTemplates (or some other authority) defines a few templates generally-accepted by the application-building community, this would address quite a bit of that.

Essentially, this is the direction we would like to go. A community-held template "library" with enough acceptance would allow application developers to interpret such data with well-defined semantics.

[gthorud]

A solution similar to what @thomast73 is considdering was published in draft form on BetterGedcom late 2011. http://bettergedcom.wikispaces.com/A+Data+Model+for+Sources+and+Citations It will be interesting to see if you are able to come up with something that will allow international transfer of citations, so my output is not cluttered with citations in sevral languages.

[thomast73]

It will be interesting to see if you are able to come up with something that will allow international transfer of citations, so my output is not cluttered with citations in English.

For output not to be cluttered with English citations implies that the citation creator provided citations in something other than English. If you receive a file from someone who creates English citations, are you expecting software to somehow migrate them into Chinese for you? Or are you expecting that citation creators will always produce two citations for the same source—one in English and one in Chinese? And if I had a list of citations associated with a given SourceDescription, do I only extract out the citations in my preferred language? And what if I find none in my preferred language but that there are some that could be displayed?

[gthorud]

Implementations of CSL are able to translate all (or most) of citations, assuming that the two lanuages use more or less the same character set. I have not extensively tested how well it works, but judging from the specification my assumption is that it works well. The ability to translate is dependent on how you choose your variables (using CSL terminology, I prefer to call them Citation Elements). The same principles as in CSL can be applied in a template solution, it is not that complex - once you have a template language it is actually relatively simple. (Most of CSL is just an advanced template language.)

I don't expect a solution to litterally translate any text, but most of the citations following a structure like Chicago or Evidence Explained can be translated - using language dependent templates. Even in cases where there is parts of the citation that cannot easily be translated, most of the citation can be translated automatically, and that would help a lot - and would be very useful both to a creator (see next paragr) or someone receiving the data.

I would not base a solution on requireing someone to have to write the meta data/citation in several languages, but I have that as an option in my spec - at the "Citation Element" (variable) level so the parts that are not easily translated can be supplieed in several languages. When rendered, the most preffered language would be chosen, if available. I don't think it is realistict to have users providing complete citations in several languages, and neither is it necessary in most cases (unless you are talking about free form footnotes with a lot of arbitrary text).

I don't know how many times I have wished I had such a solution when sending data about their ancestors to people in the US, so they did not have to complain about not understanding my citations.

Translating from English to Chinese would probably not work, but translating between English and a lot of other European languages would work.

NOTE also that you may not only be needing translation between languages, but also between styles, something that CSL can do, and something that templates could do.

Where there is a will, there is a way.

[gthorud]

I should add one important thing. When CSL "translates" citations, a lot is not realy translated, e.g. the title of something is kept in the original language. Neither are names of authors, publishers, places etc translated. Things that are translated are specific words like page, chapter, volume, April, author, edition, online, accessed etc. In order to give more flexebility in choosing "Citation Element"/variable types, I have added additional translation capabilities to be used inside Citation Elements.

[thomast73]

Ah! Now I am beginning to understand what you mean by "translate".