– DRAFT –
RDF and SPARQL Working Group

RDF & SPARQL 1.2 "What is new" stable URL

pchampin: we've been asked to provide a stable URL for our draft

<AndyS> +1

<pfps> seems reasonable

<niklasl> +1

[no objections]

<olaf> +1

pchampin: we probably need a WG decision

<pchampin> PROPOSAL: publish rdf12-new as a WG Note Draft

TallTed: we don't need +1
… for this

RESOLUTION: publish rdf12-new as a WG Note Draft

ktk asks pchampin to draft a resolution

<pchampin> https://w3c.github.io/rdf-new/spec/

ktk: does this require one or two documents?

pchampin: good question. The SPARQL document is at a very early stage, bare bones right now.

Consensus requires more than no objectsions

TallTed: if we were arguing about the content of the document that would matter, but for now we just need documents we can point to

AndyS: we could argue that there is only one document to point to

<tl> +1 to only one document

pchampin: I don't have a strong opinion right now. We could decide that all new SPARQL features belong in RDF-new

Adrian: we can defer that decision

AndyS: we need to decide right now in respect to what URLs to provide

<pchampin> note that there is currently no sparql-new spec, so we do not break any expectation in that respect

<pchampin> (on TR I mean)

TallTed: we have a good number of documents .... I doubt that anyone will only implement SPARQL. We just need to make decision and I would prefer one for SPARQL and one for RDF

<tl> i retract my "+1 to only one document" and agree with ted

<niklasl> Makes sense to me (two).

Adrian: any objections to having two URLs?

[no]
… asks pchampin to draft a resolution

<pchampin> PROPOSED: publish sparql12-new as a WG Note Draft https://w3c.github.io/sparql-new/spec/

<ktk> +1

<niklasl> +1

<gtw> +1

<TallTed> datestamped will be needed for stable *content* ... so there is small question whether they want stable *content* or stable *evolving*

<pchampin> +1

<rubensworks> +1

<Souri> +1

<AndyS> +1

<tl> +1

<doerthe> +1

<TallTed> +1

<olaf> +1

<Tpt> +

<Tpt> +1

<lisp> +1

RESOLUTION: publish sparql12-new as a WG Note Draft https://w3c.github.io/sparql-new/spec/

pchampin agrees to publish the draft accordingly

Vote about OWLification document 1

niklasl: this was a NOTE I created based upon discussions quite some time ago around expressing triple terms in respect to RDFS
… explains further details

This could become a WG Note

Adrian: this stays as a separate document, right?

niklasl: yes

tl: I am unhappy about the way it talks about literals
… it becomes like text, but triple terms are not text
… I can't subscribe to this

Adrian: for now we shouldn't talk about the content, rather creating a WG Note

tl: okay, I am not against creating a Note

<TallTed> Is the doc in a repo? or other collaborative editing space?

niklasl: there is no place in the document that says that propositions are literals. I can improve the clarity of the account though

<pfps> what document is this?

tl points to the wording he finds problematic

<niklasl> https://niklasl.github.io/rdf-struct-patterns/spec/

<TallTed> given that we're now producing a WG NOTE, can that migrate to a WG repo?

niklasl: we can discuss the wording in github issues

<doerthe> https://niklasl.github.io/rdf-struct-patterns/spec/#properties-of-formal-structures

pfps: what document and what paragraph are we talking about?

niklasl: paragraph 3 in section 3

<Zakim> TallTed, you wanted to ask that the doc go into a repo or other collaborative editing space for further refinement?

TallTed: given we've agreed that this becomes a WG Note, it should be moved to a WG repo

<pfps> I still don't see what the issue is.

pchampin: [staff contact hat on] we already have several WG Notes, and worry that we don't dilute our efforts

<doerthe> Thomas was referring to the sentence "This is not necessary for propositions and literal values, since their meaning is completely encoded by their corresponding terms—ground triple terms and literals—in the abstract syntax.", for the exact problem you need to ask him

niklasl: I sympathise with pchampin's concern
… and working first on high priority documents

<pfps> I'm still completely confused about the argument, which appears to be "X and Y are stated to share some characteristics therefore X and Y are the same".

tl: the name shouldn't be too abstract

niklasl: we can change it later as we see fit

<ktk> Create a note on "RDF 1.2 Properties of Formal Structures" (working title)

pchampin: the name isn't critical, what is important is to bring the document to a WG repo

<pchampin> PROPOSED: the WG adopts niklasl/rdf-struct-patterns for future publication as a WG Note

<ktk> +1

<niklasl> +1

<Souri> +1

<pfps> +0

<tl> +1

<pchampin> +0.5

<TallTed> +1

<lisp> +1

<AndyS> +1

<olaf> +1

<gtw> +1

<doerthe> +1

RESOLUTION: The WG adopts niklasl/rdf-struct-patterns for future publication as a WG Note

<Tpt> +0.7 (sounds like a good idea in absolute but I have not read it)

SPARQL Issue Triage 2

<pfps> Just sort the display by Repository

Adrian: I tried to find a summary but failed. Let's open the discussion.

It is unrealistic to go through all the SPARQL issues in the time we have.

It is very likely that SPARQL will become a living spec. What is needed in the short term?

AndyS: these are personal ideas: there are a few critical issues we should address (talks us through them)

These are all in the main spec

<Tpt> w3c/sparql-query#326

<gb> Issue 326 Remaining erratas and errata-like issues (by Tpt)

ThomasT: I agree with AndyS that are a few big issues to fix and some others we would like to deal with

<AndyS> w3c/sparql-query#355

<gb> Pull Request 355 Syntax scope requirements for (NOT)EXISTS (by afs)

TallTed asks AndyS to share the links to the critical issues

<AndyS> w3c/sparql-query#355

<AndyS> w3c/sparql-query#327

<gb> Pull Request 327 Clarify that unaggregated variables in SELECT raise an error (by Tpt) [spec:bug]

gtw: I agree with AndyS

olaf: I also started working on a PR relating to property terms

<olaf> w3c/sparql-query#396

<gb> Pull Request 396 Evaluation of property path patterns that contain triple patterns (by hartig)

Adrian: please make a comment

AndyS: anybody with write access can edit the document

<rubensworks> w3c/sparql-query#53

<gb> Issue 53 Incorporate RDF-star CG report (by rubensworks) [spec:substantive]

rubensworks: most of this was discussed some time ago, we can review that

Adrian: any implementor's feedback?

<TallTed> I have nothing to flag at this point

james: nothing I would object to (earlier than the last 2 weeks which I haven't looked at)

Adrian: let's use Thomas's issue as the main one to work on

<Souri> Will { ?x :p1/:p2 ?y } ever have any match if the only triple with :p1 as predicate has a triple term as object? EXAMPLE=> :bob :thinks <<( :alice :likes :bob )>> .

Adrian: we should give everyone a chance to look at this

Any Other Business (AOB), time permitting

pchampin: I wanted to inform the WG that I heard back from the TAG in respect to MIME types, they have been looking at the PR and I will chat with the staff contact for the TAG

Adrian: i've heard from the ISWC re the joint session which looks good

<gtw> is that question any different than a question about literals?

Adrian wrap the meeting up ...