COinS and Wikipedia

Iniquity
Hello! First I want to say - thank you for your work! :)

I'm one of the editors of Wikipedia and we've been using COinS for many years to provide metadata for our sources. But I did some research and noticed that starting from the official website of the COinS standard itself, ending with OpenURL - everything died. Many applications and resources also stopped working.

I have a suspicion that we need to switch to a more supported standard. Do you have any suggestions/ideas?
  • Iniquity
    We thought about microformats or pure RDF, but it’s not clear whether it will work.
  • DWL-SDCA
    As it has been a couple of days without a reply from someone authoritative, I'll ask a clarification question.

    Are you wondering about allowing Wikipedia users to one-click cite articles that appear on the Wikipedia website? I can already one-click Wikipedia article references into Zotero (but without the article title).

    I use unAPI with my website and Zotero captures the records perfectly. However, I've been getting the impression from some of my technology reading that unAPI is no longer among the best-supported formats.

    unAPI was a simple process to implement but I wonder if the value of that system is fading.

    Maybe you were asking a different question. Maybe experts here better understood your needs. I apologize for semi-hijacking this thread with a question about unAPI if that is far from the topic.

  • Iniquity
    edited March 25, 2024
    Thanks for answer! Yes, after some research I also noticed that the unAPI is not used much. And as far as I can see, it is not even semantic, since it uses abbr for definitions. I could be wrong.

    > Are you wondering about allowing Wikipedia users to one-click cite articles that appear on the Wikipedia website? I can already one-click Wikipedia article references into Zotero (but without the article title).
    I'm worried that the metadata of the sources (references) we use may no longer be supported.
  • adamsmith
    There's some ongoing work on generic translation on Zotero that will support JSON-LD embedded in a script tag anywhere on a website, so that'd be a very reliable approach, but might sit awkwardly with Wikipedia/Wiki.

    Beyond that, I think some sort of microformats (also schema.org based, probably?) are probably the best bet, but Zotero currently doesn't have any support for them and they're also a bit tricky to support for bibliographic metadata.
  • Iniquity
    Thanks for the answer!
    >There's some ongoing work on generic translation on Zotero that will support JSON-LD embedded in a script tag anywhere on a website
    Yes, it seems to me that the script tag is a little not about Wikipedia :)

    > (also schema.org based, probably?)
    schema.org has some problems, you can read here: https://en.wikipedia.org/wiki/Help_talk:Citation_Style_1#c-Trappist_the_monk-20240323234000-Iniquity-20240323220500
  • adamsmith
    Yes, it seems to me that the script tag is a little not about Wikipedia :)
    I mean -- it's not obviously so. There's not a super stronger reason for Wikipedia not to insert a script tag with non-displayed metadata instead of a span tag with non-displayed metadata
    schema.org has some problems
    How are you thinking of the metadata on the pages? My understanding (and certainly the current function of the COinS) would be that it's the metadata of the cited work, not of the citation (you can't import citations into Zotero anyway, say, only the work that's being cited) and so pagination for books doesn't make much sense as a requirement (people cite a specific page, but the cited work is the book).

    If you're looking for something more ambitious that fully captures the details of the citation -- I don't believe that exists, so Wikipedia would have to innovate that, but I'm also not entirely sure of the use case of making the cited location machine readable.
Sign In or Register to comment.