Meeting minutes
Approval of last week’s minutes: 1
ora: any comment?
pfps: there is a duplication, but that's not a big issue
<ora> PROPOSAL: Approve last week's minutes
<ora> +1
<pfps> +1
<AndyS> +1
<pchampin> +0 (was not here)
<ktk> +1
<tl> +1
<niklasl> +1
<Enrico> +0 (not there)
<gtw> +0
<TallTed> +1
<lisp> +1
<olaf> +1
<fsasaki> +1
RESOLUTION: Approve last week's minutes
relation with SHACL
ora: AndyS drew my attention to the fact that the SHACL WG are ready to go to CR.
… They are depending on our specs.
… Is this a problem that they go to CR before us?
… Note that we took our time for good reason.
TallTed: the SHACL WG is really eager to release early.
… I think it should be able to publish for RDF 1.1, and sync with RDF 1.2 later.
AndyS: the SHACL WG has 2 phases.
… The first one is to update SHACL-Core.
… Was planned for this summer, but well... :)
… It includes static constraints for RDF 1.2 features, that is Triple Terms and reification.
<fsasaki> as addition to what Andy says, see e.g. https://
AndyS: There should not be any process issue. Normative references can be one step before or ahead.
… It is more about the messaging.
Identifying issues to solve before CR 2
pfps: a quick one: RDF Concept changed some of its anchors, this breaks links in Semantics
… I can change them in Semantics, but I want to be sure they won't change back.
<pfps> I'll try to fix them in Semantics
AndyS: I believe that broken anchors in Concepts are fixed. In any case, we do not plan to change any.
pfps: Some links are still broken. Ok, I'll make the changes then.
AndyS: I think we should consider which Horizontal Reviews we can start now, even if we don't feel ready for the TAG.
<pchampin> +1 AndyS
pchampin: think there is some self-assessment to do ("explainers?")
… we are at the point where some horizontal reviews can start
ora: want the group to agree that the chairs can start the process
… anyone have any concerns?
<lisp> +1
<niklasl> +1
<Enrico> +1
<olaf> +1
<tl> +1
ora: accessibiility, internationalization, secruity
ora: I'd like the group's blessing to the chairs kick this off (Security, I18N, Accessibility) this before the next meeting. Any objection?
<AndyS> +1
<pfps> fine by me
Issue 228 RDF Concepts section ordering (by afs) [ms:CR]
<gb> CLOSED Action 4 decide on the short names of the specifications (on pchampin) due 22 Dec 2022
AndyS: I propose to remove the ms:CR label. I don't think that a reordering of the section would invalidate the Horizontal Review.
<TallTed> reordering is editorial. and doesn't invalidate horizontal review.
ora: agreed
<Zakim> pfps, you wanted to ask about private email
Pat Hayes reactions
pfps: what about the private emails mentioned last week?
ora: I sent an email to Pat asking if I could share his email with the group.
… He has not answered yet.
PR on "abstract syntax" vocabulary
lisp: I made a PR to fix the "abstract syntax" vocabulary in RDF Concepts
… niklasl has reacted to it at length
… what should I do with his suggestions? Should I open a new PR?
TallTed: integrating them in the current PR would be better.
<gb> Pull Request 232 revise to use "abstract data model" to unify "abstract syntax" and "data model" (by lisp)
<lisp> the #129 pull request is : w3c/
<gb> CLOSED Action 129 write a PR on rdf-concepts for the unstar mapping (on pchampin) due 2024-10-01
Move datatypes definitions to RDF-Schema?
pchampin: w3c/
<gb> Issue 163 Move datatypes definitions to RDF-Schema? (by pchampin) [ms:CR] [spec:editorial]
pchampin: need not be blocking so remove ms:CR (or close)
ora: prefer to close - can reopen
niklasl: I'm for postponing, I like the direction of it.
<tl> +1 to niklas
ora: nobody insisting to do it now, let's postpone
pchampin: done
w3c/rdf-star-wg#169 and w3c/rdf-star-wg#128
pfps: unless new information is available, I'm again proposing to close w3c/rdf-star-wg#169 and w3c/rdf-star-wg#128 without any change to RDF-Semantics
<gb> Issue 169 definition of reifiers is non-normative and seems vague (by rat10) [needs discussion] [propose closing] [ms:CR]
<gb> Issue 128 map the annotation syntax to `rdfs:states` (by rat10) [enhancement] [propose closing] [ms:CR]
tl: I have PRs open to address this, which do not add anything normative, but contain informative text that I believe should be there
… We discussed last week whether we need a model-theoretic semantics for statements about statements.
… Nobody has proposed any, and I can't propose it, so this seems moot, though I would prefer if we had it.
lisp: following last week discussion, I wrote a note arguing that the current text is not sufficient for readers / implementers of the spec.
… I can not support the current document without a formal semantics for statements about statements, or an explanations why there should not be one.
Enrico: if I understand pfps's answer, the current version of the spec are able to talk about statements about asserted statements.
… I explicitly replied to Thomas, so I don't see why we are still discussing it.
… Granted, there is not a single construct to express this semantics, but the current semantics supports it nonetheless.
pfps: IFAICT there are two issues there.
… one is whether the semantics machinery should be changed.
… I agree with Enrico that the current state goes as far as it should go.
… Going further is nor for RDF, maybe N3.
… the other issue: there was a section in Semantics about the vacuity of old-style reification.
… It didn't talk about semantics, it talked about vacuity.
… It has moved to RDF-Schema, which I believe is a better place for it.
… Description of the intended meaning of these constructs should be in Concepts or Schema, not in Semantics.
… And Concepts already contains a long section describing this intended meaning.
lisp: it is consistent with the semantics, then it should be in Semantics.
pfps: I completely disagree. There are many other cases in RDF that are not in Semantics, e.g. rdf:first .
lisp: this is different. rdf:first is a single statement.
… reification introduces a relationship between several triples (the reifiing statement and the statement about the reifier)
… this is something completely new in RDF, and should be described.
pfps: I disagree with this reasoning.
… RDF 1.2 introduces triple terms, and RDF 1.2 Semantics describes what they denote. That's all that is required.
… The rest is about the intended meaning of the reifier, which does not belong to Semantics.
lisp: the explanations I got in emails are satisfying, but I don't see them in Semantics, and that's where I want them to be.
<pfps> what "concise statement"?
<tl> pfps: w3c/
<gb> Pull Request 220 Annotations on assserted triples are based on operational semantics (by rat10) [ms:CR]
TallTed: as I understand, you are not looking for the text answering your question, you are looking for its place in the document.
… The editors of the document disagree with you on this, but if the group agrees to put this text in this documents, the editors should apply the group's decision.
Enrico: Section 5.3 of the Semantics documents describes all the connections about triple terms object of a reifying triple and asserted triples.
… The semantics is described, and the text is there.
<Zakim> TallTed, you wanted to add one thing
pfps: the discussion is about a concise statement that I'm supposed to have authored but I still don't know what this concise statement is.
TallTed: there is the question of whether the disagreement is bad enough that it will raised to formal objection.
<niklasl> It is in this comment AFAICS: w3c/
<gb> Pull Request 220 Annotations on assserted triples are based on operational semantics (by rat10) [ms:CR]
TallTed: this discussion shows that some clarification is needed.
<tl> +1 to ted
<Enrico> I would like to remind the text in Section 5.3 of semantics <https://
TallTed: lisp, it would be helpful if you put this concise statement in a document so that we can collectively decide as a WG whether it belongs there
lisp: pfps, I will retrace this concise statement in emails.
… Enrico, is the text you quote above supposed to be the sufficient explanation?
Enrico: yes, the Semantics document is about definitions, not explaining how to use the knowledge graph.
… We may want to develop these definitions in natural language somewhere, make your proposal and we will see.
niklasl: I pasted a link to what I believe is the concise statement that was mentioned.
… As pfps hinted, it is a reformulation of what's already there.
<niklasl> "The precise statement" is a slight reformulation of what is already there (set of propositions and the set of facts). More crucially, does the precise statement change the requirement of two triples (a reifying one and the one asserting the triple term of the reifying one)?
<niklasl> I.e. would "IEXT as a subset of the set of triples" still not require asserted triples to be present in the graph for t to be in IEXT?
ora: I want to put this issue behind us, we need to vote on this.
<niklasl> T. from "RE(a,IS(rdf:reifies),t) is in IEXT, and t is in IEXT, for t = RE(I(x),I(y),I(z)))." of the "precise statement".
<pfps> There are several parts of the comment - one is a *change* to the semantics and one is a description of how the *current* semantics works and how reification could be described (elsewhere)
ora: we can't discuss this forever. Some argue that what we have is sufficient, some disagree.
<niklasl> Which i think is a reformulation of the set of facts being a subset of the set of propositions.
<Zakim> AndyS, you wanted to suggest that concepts/128 can be closed. This has been suggested -- May 22
ora: It seems that we can only solve this by voting, but we need to understand what we are voting on.
AndyS: we started this discussion on 2 items. I believe that we can close w3c/rdf-star-wg#128, the discussion is focused on the 1st one.
tl: I agree to close it. It would have required this discussion to be solved, but we are not there yet.
ora: it seems that we have 3 options: not include this new test, include it in Semantics, include it somewhere else. The Primer would sound like a valid alternative.
lisp: I would like to ask the editors where they advice to include the concise statements pointed to by niklasl.
pfps: what niklasl pointed to has 4 paragraphs, the 1st of which is a change to the current semantics.
… Is that the concise statements we are talking about?
<lisp> this is the statment The way to go would be to define the set of triples in an interpretation as the range (the mathematical range not the RDFS range) of RE and define IEXT as a subset of the set of triples. Then an asserted triple reifier of a triple term, x y z, in an interpretation is just a domain element, a, for which RE(a,IS(rdf:reifies),t)
<lisp> is in IEXT and t is in IEXT, for t = RE(I(x),I(y),I(z))).
pfps: I believe that the change would damage the semantics.
<niklasl> Would something like "A resource relating to a proposition that is in the set of facts thus relates to a fact." do as an addition in Semantics 5.3?
Enrico: is the issue that you have problem understanding 5.3, or that you think people will not understand it?
lisp: my problem is that 5.3 does not describe what rdf:reifies means.
Enrico: that's not what 5.3 is about. We should take this discussion over email.
TallTed: if possible, the output of this conversation is likely to be an actual short statement which James feels is appropriate to add somehwere in the documents that exists. A PR doing that would help us evaluate that statement and whether that's the best place to do it. If we have it by Wednesday or earlier, we can discuss this in next week's
meeting.