Meeting minutes
Approval of minutes from the last two weeks:
Erich_Bremer: I work in bioinformatics with pipelines.
… I've been working with RDF and RDF-star, which will work nice
Erich_Bremer: I was briefly in another group, but this is pretty much my first WG.
ora: We look forward to hearing your real-world use cases.
pfps: I'm looking at the meeting announcement, and it's missing the links.
ktk: I forgot to change that we don't put scribes as the first item anymore and things changed.
<ktk> https://
<pfps> the email announcement is fine. it's the description on w3.org that is missing the links
Approval of minutes from 2023-10-19
<ktk> pfps: I updated it now
<pfps> minutes look fine to me
<gkellogg> +1
<AndyS> +1
<Dominik_T> +1
<tl> +1
<AZ> +1
<ora> +1
<olaf> +1
RESOLUTION: approve minutes of October 19
<ktk> https://
Approval of minutes from 2023-10-26
<pfps> last week's minutes also look fine to me because the presentations have slide backup on the email list
<ktk> PROPOSAL: approve minutes of October 26
tl: There was a correction on the abstract syntax (which was captured)
TallTed: You can use SED substitution to correct the log.
<ktk> +1
<pfps> +1
<gtw> +1
<AndyS> +1
<ora> +1
<gkellogg> +1
<olaf> +1
<tl> +1
<Dominik_T> +1
<Souri> +1
AZ: What do we do with [to be completed ...]
… I think we missed something niklasl said.
gkellogg: it was about the need to use the Abstract Syntax and not higher-level syntactic sugar
ktk: niklasl isn't hear now.
TallTed: Using ??? rather than "to be completed" makes it easier to just accept.
<AZ> +1
RESOLUTION: Approval of minutes from 2023-10-26, with corrections by pchampin (noted down by ktk)
Proposal for next week's discussion
<ktk> https://
ora: We need to come up with a topic for next week.
… We had a lot of good discussion with many proposals, but we need to decide what we're going to pick up.
… We need to make a decision on what to do, but some more discussion is still needed.
AndyS: We also said to look at the use cases to see how they fit the proposals.
… They're quite big, with a lot of detail. We need to get down to a few examples that capture the range of needs.
… This could replace the "seminal example".
tl: I agree; I started trying to do that, but I'm at ISWC next week.
… I've been looking at the Abstract Syntax, and think that my proposal can work without changes. It looks just like named graphs, but requires some changes to how we model.
… That will take some time to work out.
ora: anyone else at ISWC? (olaf and AZ)
tl: niklasl is also not here next week.
ora: We're loosing some significant participation.
… Should we have the meeting or postpone? That would make the next long meeting in three weeks.
Souri: One of the upcoming Thursdays (Nov 23) is Thanksgiving, which is a holiday in the US.
… Nov 23.
AndyS: How about we adjust the schedule?
Souri: We could convert a 1-hour meeting to a shorter discussion.
ora: Why don't we have mainly discussion for the next couple of meetings.
… Cancel next week, and have a 1-hour discussion in two weeks.
ktk: The next long meeting would be Dec 7th.
ora: that will allow us to make some progress.
gkellogg: I'm concerned about how much time we're loosing
ora: If someone would like to schedule something longer, I'm happy to attend and help.
… Next week is shot due to ISWC. Could we have a longer meeting on Nov 16th?
… Maybe we could shift the time?
Souri: If we could go an hour later, that would help.
… If we could go 12-2 instead of 11-1 that would help.
… Also, US goes off DST this weekend.
ora: I'm happy to do that.
ktk: The main problem for me on the 16th is I'm in Jakarta.
… I would be unlikely to make a two hour meeting.
ktk: We'll do a two-hour meeting on the 16th starting an hour later.
ktk: the next long one would be Dec 7, then Dec 21 then Jan 4th.
enrico: I can't join a later meeting. What about the semantics groups meetings? I think we should delay until we have more consensus for our direction.
ora: the TF is free to decide it's own schedule.
AndyS: For the meeting on the 16th, what inputs should we bring. We need a way to have discussions that converge.
… There's a lot of material, but maybe we can reduce to a set of facts to foster convergence.
ora: We have material, so I expect at a minimum people to decide on a proposal they support.
… I wouldn't object if someone wanted to synthesize something from existing proposals.
tl: I think we should focus on a use case and use that to measure the different proposals.
… But, we need to focus on what that is.
… AndyS proposed to have a simple use case example to focus on.
AndyS: I didn't say a single use case, we could have a couple. For example the Superman example.
… The point is that examples should be a couple of lines long that have the essence of the problem.
AndyS: pfps, what's your feeling of what use case makes more engagement.
… There's still a lot to go through, and it's pretty long.
pfps: A use case document is often longer than what's on the website.
… At one time I created a summary page that lays them out, and it's still up to date.
… They don't have examples, but they are distilled.
<pfps> The Summary is at https://
pfps: That only has completed use cases, there are a couple of more that are in process.
… I can update that.
… I can add some RDF to the examples, as well.
ACTION: pfps to update Summary wiki page with new use cases and RDF snippets
<gb> I created issue #8
<gb> but I could not add the "action" label.
<gb> That probably means I don't have push permission on w3c/rdf-common.
ktk: For Erich_Bremer, I think it would be good for you to go through the use cases to see if there's something you think is not covered.
close #8
<gb> Closed issue #8
<tl> Is Enrico's question about sem TF meeting still up?
ora: Between looking at use cases and proposing examples, is there something else?
AndyS: I think there's more to do, but not realistic right now.
Review of open actions, available at
<ktk> Open Actions: https://
ora: No actions to discuss without pchampin.
Review of pull requests, available at
<ktk> Pull requests: https://
gkellogg: I think the directional language-tagged string requests can be merged
AndyS: What do we need to do about rdf:JSON
gkellogg: not sure how to resolve without a concensus.
… Mostly about I-JSON limitations.
pfps: I have two concerns: one is that JSON is JSON and has a syntax. It is not limited to IEEE floating-point syntax, and string syntax is not limited to nice codepoints.
… If you're going to say JSON, do JSON.
… Otherwise, we'll confuse everyong.
… Also, I-JSON is not well defined; I don't know how to know what is I-JSON and what is not.
… It has to do with what floating point numeral relates with IEEE.
… If we're going to include it, someone's going to have to come up with a good definintion.
ora: Let's say we go with what JSON says, do we need some guidance?
gkellogg: Codifying the work from JSON-LD
… uses canonical form
… it is what happens in JSON-LD libraries
… has been used for several years
… there is now an RFC - wasn't at the time.
… JSON-LD processing duplicates what is now the RFC
<TallTed> JSON Canonicalization Scheme (JCS) -- https://
https://
https://
gkellogg: JCS is mentioned
pfps: This is not I-JSON?
… defining doc from JSON-LD is 10.2 - lex space is all of JSON.
gkellogg: JCS used to get to value space
pfps: lex to value is parsing consistent ECMAscript.
gkellogg: my impl uses a different JCS impl.
pfps: what is the lexical space?
gkellogg: effectively I-JSON because only way out of JSON-LD.
gekllogg: JSON-LD is RDF with constrainsts like no surrogates
pfps: does not limit lexical space to I-JSON.
ora: I'd like to see pfps and gkellogg work on a solution that is practical.
<TallTed> perhaps pfps and gkellogg could move this to the mailing list, or have a separate zoom session (perhaps using semantics TF call time), that others could also join
ora: this is an important issue to get right.
ora: Next week is canceled.
AndyS: There's an issue for deciding the criteria for EXISTS.
<TallTed> gkellogg, pfps -- if you set up a special call, please have ora or AZ or pchampin add it to the group calendar, so it automagically shows up for subscribers. (Knowing my weaknesses, I will likely overlook it if announced on the mailing list.)