Topic on User talk:TBurmeister (WMF)

Xover (talkcontribs)

Hi TBurmeister,

I saw the new people in Developer Advocacy listed in the Technical Community Newsletter and, always happy to see this area strengthened at the WMF, took a look at your user page. Going by that you're going to be my new favourite person at the WMF! 😎

I currently work mainly on English Wikisource (transcribing old public domain books, like Project Gutenberg but in a wiki-align way), so bibliographic metadata and bibliography is bread and butter for us. We also try to delegate as much as possible of such data to Wikidata, by adopting a common information model for bibliographic data (work—edition-division being the core of it). Our transcribed texts are also backed to scans of a physical copy of the book, which are hosted on Commons (unless a policy difference between the projects force us to host them locally on enWS).

Wikidata is one approach to structured data, but with their own policies and priorities. Commons has the "Structured Data for Commons" initiative, which is a different initiative, with different goals and priorities, subject to different policies. Wikisource has no real software-backed structured data, but we try to be structured about the way we store bibliographic data. For all of these, one major goal is to be able to automatically link to a relevant article on Wikipedia (about the author, or about the work; rarely, but not never, to link to an article about a topic): which means Wikipedia's ontological approach to article scope and disambiguation ends up being a relevant factor for the information model on Wikisource.

As you can imagine this setup is quite complicated for even experienced and technically inclined contributors to deal with, and essentially impossible for new contributors. We desperately need better software support here, of course, but there's also a pretty huge gap in good written documentation for how this works at various levels (information architecture, and information models, being two such). So… If the scope of your duties puts you anywhere near this little nexus of challenges, it's a problem domain that I think you might find interesting; that you might be uniquely qualified to tackle; and that could potentially have an outsized impact both short-term and long-term.

In any case: Welcome to the Movement! (or at least to the new job; your user page doesn't mention on-wiki experience so I'm assuming you're new to the movement)

PS. There's also a years-long and strangely half-baked effort to improve bibliographic and citation metadata and better support reference use on Wikipedia that seems to mostly have generated discussions at various conferences and nothing much tangible. There's a massive potential for outright replacing Worldcat and improving global bibliographic metadata, huge potential for GLAM collaboration, common cause with the Internet Archive and other non-profits, etc. I don't think anything's happening there, but if it was it might be an effort right up your alley, so you might want to look into it (I can try to find you links if you like, but I don't have them immediately to hand).

TBurmeister (WMF) (talkcontribs)

Hi Xover! Thanks for the welcome! You are correct that all this stuff is very up my alley :-) I'm not entirely new to the Movement, but have been dormant after I organized an edit-a-thon with some fellow librarians back in 2016. (I added a link to my volunteer account on my user page, thanks for pointing that out).

I wasn't aware that Wikidata and Commons had different approaches to structured data, nor that Wikisource lacks one. Interesting. I wonder if the effort you mention in your last paragraph is the 1lib1ref initiative?

Anyways, in terms of the scope of my staff role: I'm focused on documentation for technical contributors, which includes data contributors or data re-users. Note, though, that User:Dan_Shick_(WMDE) mostly manages Wikidata documentation. My role doesn't involve building data models for the large-scale, project-level structured data initiatives you mention, but it **does** involve documentation to support developers who are trying to use or apply those data models. And, most importantly, it involves providing tools and education to support and empower people to improve the technical documentation that matters to them! I'd be interested to hear more about the documentation you use and what types of technical tasks you think need more doc support, and I can collaborate with you or other members of the community to plan for and implement those improvements if that's something people want to do. Also, if you're aware of any existing Phabricator tasks that you think capture some of the pain points, feel free to link me to those.

Nice to meet you, and thanks for reaching out!

Reply to "Welcome to the movement"