Meeting minutes
<AndyS> Anyone know where the scribes list is kept? https://
<ktk> AndyS, I rotate them in my head
<ktk> I do go through the one you posted I think https://
Approval of minutes from the last two meetings:
https://
PROPOSAL: Approve minutes 2023-11-30
<gkellogg> +1
<niklasl> +1
<ora> +1
<TallTed> it needs a s/metings/meetings/g :-)
<AndyS> +1
<tl> +1
<ktk> +1
<TallTed> +1
<AZ> +0 (I was not there)
<pchampin> +1
<doerthe> +1
<enrico> +1
RESOLUTION: Approve minutes 2023-11-30
https://
PROPOSAL: Approve minutes 2023-12-07
<niklasl> +1
<gkellogg> +1
<AndyS> +1
<tl> +1
<pchampin> +1
<ora> +1
<doerthe> +1
<ktk> +1
<TallTed> +1
<AZ> +0 (I wasn't there)
<eBremer> +1
<pfps> can pierre-antoine put a topic into these minutes?
<pfps> ... topic suggestion to follow
RESOLUTION: Approve minutes 2023-12-07
Proposal for next week's discussion
<pfps> ... probably something like "discussion on how to proceed"
Ora: We did not converge on anything. There are a lot of interesting proposals on the table. As the chair, I wouldn't want to kill them off but the reality is we need to come to an agreement about the direction of our work. Lately there was more about Souris proposal. I'm not done with my analysis.
… It looks interesting.
… I understand the goals of it.
… I guess regarding that my question is about the existing RDF 1.1 semantics.
AndyS: Are we having a meeting next week? It's close to Xmas. And the next few weeks.
<TallTed> pchampin - I've messed up the pre-item-1 section beyond `s` fixing. please clean that after meeting ends
Souri: Please look at my slides I posted today at 4AM, not at the ones before.
<pchampin> I can attend next week, but I may not be available on 4 Jan
<Souri> +1 next week okay for me
ora: Proposal is we keep the meetings on 2023-12-21 and 2024-01-04
<olaf> I wouldn't be available for any of the two meetings (neither next Thursday nor the Thursday after)
<enrico> I can attend next week
ora: and only the one in between is cancelled
<doerthe> I also can't attend both meetings
<pfps> 21 is fine by me
Ora: We cancel the one on 2023-12-28.
Ora: So what are we gonna talk about.
tl: To understand the timeline. The WG is scheduled until August. At what time would we have to publish something. What is the timeline for a decision?
pchampin: The steps for getting a recommendation is a proposed recommendation, which needs to be approved by the members. Before that we need a candidate recommendation. And we at least need 28 days for that. So in total at least 2 months, which is a very optimistic, non realistic scenario.
<AndyS> https://
pchampin: We are late based on our planing. Realisticly we will need a charter extension.
gkellogg: A new charter needs to reflect the reality. But most important it needs to reflect what we want to do.
<TallTed> fwiw -- Souri's latest (4am) RDFn slides -- https://
ora: I would like to understand Souris proposal, with the RDF 1.1 semantics compatibility.
<pfps> +1 to considering use cases
niklasl: Use cases. Not opposed to Souris proposal, pfps analyzed them and we need to keep them in mind. For this proposal and possibly some others.
… Also the proposals regarding the doubts of named graphs. They should be kept in mind in the background.
pchampin: +1 on use cases, I would like to see how it impacts the abstract syntax as well.
… I know Souri said it's not implementation focussed, but it still has very much an implementation taste for me. So I'm interested what changes it applies to the abstract syntax.
ktk: we has a discussion with Ora about the impact of the different proposals,
<niklasl> +1 for meaning for abstract syntax (e.g. what "multiple asserted" within the same graph means)
ktk: we are taking about a 1.x version, we are not supposed to break things.
<TallTed> +1 semantic-style versioning. 1.2 should not break 1.1, 1.0. 2.0 branding will be needed for breaking changes.
ktk: I would like to look at each proposal in that perspective.
<AndyS> https://
AndyS: I tried to summarize my take of the discussions and put them on the list.
… I would like to talk about that.
tl: RDF Star introduces a new term, this doesn't break RDF 1.x but it is a fundamental change.
ora: if we say it doesn't break it means the existing RDF 1.1 continues to work as it has.
… can the next week's discussion be use to discuss Andys summary proposal and Souris work
enrico: I was thinking about Andys proposal. This merits some discussions, it's a simple and clear proposal.
<TallTed> +1 discuss AndyS's post [1] and Souri's slides [2]. [1] https://
enrico: The minimal thing is to do what was proposed in the CG. Or we can go further like Andys proposal.
… We can discuss this further in the semantics TF tomorrow.
Ora: What TallTed put in IRC is a very good description of the topics.
Review of open actions, available at
https://
pchampin: Shortnames now point to the latest REC, not the very latest.
… Some missmatches are made consistent as well.
gkellogg: n-triples does not seem to point to the right one yet.
pchampin: Superseding is still a mistery to me. I keep investigating.
<Zakim> gkellogg, you wanted to ask what about https://
<Zakim> pfps, you wanted to claim victory on 4/
<niklasl> +1 for "no less than CR"
pfps: I'm claiming victory on mine. I looked at all use cases from the summary page.
AndyS: I claim victory on mine.
Review of pull requests, available at
https://
<Zakim> AZ, you wanted to talk about PR on scribes.md not listed here
<TallTed> w3c/
AZ: I have a pull-request on scribes but it does not show up
ktk: I will take care of it, thanks.
ACTION: pchampin to add the rdf-star-wg repo to the dashboard
<gb> Created action #100
pchampin: I will add the RDF-Star WG repo to the dashboard
<TallTed> there are a few more languishing PRs on rdf-star-wg -- https://
ora: The JSON one is still open.
gkellogg: We reached an impasse
pfps: someone will need to nail down the lexical space.
<niklasl> +1 for until more bandwidth
gkellogg: I suggest we defer this issue until we come to conclusion of the more fundamential issues.
ora: agreed.
Issue Triage, available at
https://
gkellogg: we better discuss https://
… regarding that, once we finished case normalization & concepts, that brings back what do we do when normalizing n-quads and n-triples.
AndyS: yes, I thought we agreed on that and not mentioning particular case concepts in RDF concepts. Because it goes into canonicalization
gkellogg: I will do a pull-request based upon that direction.
<gkellogg> https://
AndyS: implicit timezones. For SPARQL or the web, implicit timezones are less useful. Proposal to make this explicit.
pchampin: I sympathise with Andys reasoning. But setting it to Z makes it more likely to betray the authors intention.
AndyS: that's the problem with deciding, someone will be unhappy.
<pchampin> In the face of ambiguity, refuse the temptation to guess. https://
<niklasl> I don't *want* to refuse to guess, but I think that is the more careful (perhaps even more responsible) way
<gtw> w3c has a mastodon server
<gtw> w3c.social
ora: W3C annouced to leave Twitter/X.
<TallTed> https://
ora: maybe we should have an account for our group at some point.
<AndyS> niklas - when sorting a mixture of with and without timezone, it's a choice point (undef comparisons go in a different place in the ordering)
<AndyS> I've linked the discussion here and removed "needs discussion" label for now.