System, Stian Soiland-Reyes, Phil Archer, Tom Demeranville
None received
Stian-Soiland-Reyes
Meeting began at 11:01:22 | |
scribe | Stian-Soiland-Reyes |
hello | guys |
shall we start? | |
ok | we'll wait a few more minutes for Phil |
Phil Archer | OK I'm all set |
Stian | Phil mainly pointed out some issues on URIs. Phil: URIs identify different things. Person is not bytes. Person is not updated last week! So that's why we need to know what we are talking about. The actual person, an account, or the cat they own, etc. All different URIs. What Sian did originally was to make the differentiation the person and account by adding a / to the end of the URL. The problem of that it is easily confused - generally the browser does this. I stood up in front of academic people and said this was terrible! And then I had a look at it.. and see.. no, it's not the same! Just one character difference. I was a Research Data Alliance and talked to Laura, and we talked about this. So I said I would offer some suggestions, which ORCID would either ignore or take on board. That was a long time ago! http://www.w3.org/2015/03/orcid-semantics |
Tom | Talking to this with the dev team.. it's easy enough to add such changes to the implementations |
… hoping we would get it right. But we're OK with doing whatever you recommend. | |
Phil | Problem is that any change must be.. Do we have any sense of existing users of the data? Tom: Stian: Like we have been using ORCIDs in our RDF, but without programmatically accessing the info (but relying on the ability to do so) Tom: We have some other version of the API.. and RDF can get a bit outside the versioning here. Phil: Also this is a bit secret.. no indication anywhere that this existed. Stian: Like a button somewhere Phil: Like I suggest how to improve navigation. Tom: Redirection is already part of the API.. but it is still marked experimental and not part of the documentation. So improving this description is one of the tasks we have there. |
Tom | As it works differently from the other APIs it is a bit different from the REST APIs |
… But it will be become part of the API | |
… Adding extensions is a bit tricky.. like we have more than one .json representation. But we could do like .ttl | |
Stian | If it's .ttl or /experimental-rdf or so is not as important.. Tom: Will end up on one of our API endpoints. Could you clarify something? |
Tom Demeranville | http://pub.orcid.org/v1.2/0000-0000-0000-0000 ">http://orcid.org/0000-0000-0000-0000 should 303 to http://pub.orcid.org/v1.2/0000-0000-0000-0000 |
So the first URI is the person, the second is the profile document | |
this solves the / issue | |
Phil | One worry here with the second URL is the version number.. |
Tom | Our API is versioned.. |
… it's not critical to get rid of the version number, the big thing is the redirect | |
Tom Demeranville | curl -i -H 'Accept: application/rdf+xml' https://orcid.org/0000-0003-4654-1403 |
Serif? | |
Phil Archer | CERIF |
different URI structure needed for CERIF as it's a separate API | |
Only person info is returned in the ORCID | |
what about publications | |
Stian can't do: foaf:publications as then the person is also a page (foaf:publication range is a foaf:Document) but perhaps foaf:publications with corresponding id="publications" in the HTML | |
Phil Archer | http://xmlns.com/foaf/spec/#term_publications |
Stian | IDeally also publications could be in RDF,, but more work would be needed in choosing ontologies etc. here |
Phil | My concern is to push you in the right direction |
At W3C in our spec-writing tool we are trying to make it easy to include people's ORCID | |
Stian | with PROV statements of the spec? |
Phil | One of the requirements we found internally.. consensus was that requirements of linking to a profile, it must return RDF (among other things) |
Stian Soiland-Reyes | http://fantasai.inkedblade.net/style/design/w3c-restyle/2016/sample |
? | |
Phil Archer | Example of W3C usage of ORCID http://w3c.github.io/dpub-pwp/ |
Phil | I'm talking from a theoretical point of view, while you are on the practical side. So it has to work for ORCID's users |
… Just get the basics right.. then the community might ask for more :) | |
Tom Demeranville | Summary Change to 303 redirects Change foaf:publications to {html url}#activities Change personalProfileDocument to {api url} |
Tom Demeranville | Add documentation |
303 redirect fix: https://github.com/ORCID/ORCID-Source/pull/2472/files | |
Phil | ... crossref etc.. should be semantically aligned |
… Also talking to Ms Yan on CERIF in another projects.. Very small world! :) | |
Tom | Get some basic Dublin Core |
Stian | Will try to improve the Jena usage in the code so it's easier to expand vocabulary use - no need to use the ontology support in Jena really. |
… You also mentioned provenance? | |
Tom | Yes, about activities.. DOI relationships |
e.g. a record can be imported several ways | |
tricky for us. Early on we didn't distinction between how it is added. | |
if datacite add something to datacite.. is that because you did it, or because a datacite bot did it | |
Phil | sit on UK Research Data Forum. Nick Right from NEwcastle. |
… Lots of what we talk about here.. orcid, crossref, datacite.. getting the semantics usable, is relevant to that group | |
Timelines | Tom is trying to get it into testing by end of year. |
as changes are quite minor | |
Phil | Hoping a small delta would have some impact :) |
Ivan Herman is on some advisory boards for ORCID | |
both Phil and Ivan are happy to poke around and review | |
Phil Archer | http://philarcher.org/tools/skypescribe/ |
Stian Soiland-Reyes | http://s11.no/s/ |
System | skype ?chat&blob=2sQ3HkRVO04dDjG8jS2ayP6REi2Yj_iRp04CMjRdNI4tqBhxFqGp7ZTgePxA4yBT_0wDrYgfXi-LOMl9 |
Stian Soiland-Reyes | http://s11.no/s/?chat&blob=2sQ3HkRVO04dDjG8jS2ayP6REi2Yj_iRp04CMjRdNI4tqBhxFqGp7ZTgePxA4yBT_0wDrYgfXi-LOMl9 |
ACTION: Tom to review the 303 pull request and put into QA | |
ACTION: Stian to simplify Jena code usage so it's easier to modify vocabularies | |
ACTION: Tom to Modify URIs used for account and profile document | |
ACTION: Tom to Keep Phil/Stian/Ivan updated on RDF progress (due end of 2015) | |
Meeting ended at 11:42:11 |