Meeting minutes
<AndyS> https://
Task force reports
HolgerK: No updates for Core or Node Expressions. Changes to SPARQL for the rules.
… Main changes: introduction of iterations. SHACL-AF only had 1 loop, now we have a good design for iterations aligned with SRL, with layers produced by stratification or set manually. Order is there but no longer really important.
… Flag for run-once rules, they may e.g. produce blank nodes etc.
… Support for global rules has been added, not all rules have to be attached to shapes anymore. This is also aligned with SRL.
nicholascar: For Profiling: PRs have been addressed, for background set-up tasks, including definitions. Needs another pass.
… Profiles vocabulary is not yet out as FPWD.
… SHACL profiling meeting this week, the plan is to tidy up documents and find incomplete things, start getting close to a final doc.
… Profiles and recursion need more work.
ieben: For SHACL UI: SHACL renders discussed, e.g. for cases with no focus nodes, not all complex scenarios will be covered.
… config resolution for components, auto-complete discussed by Daniel.
… The issue below (1083) needs more attention and the TF is having a hard time coming to an agreement, feedback from the rest of the group is requested and is welcome in the issue itself.
<ieben> w3c/
nicholascar: How close are the TFs to finishing off their deliverables?
… Core is in a good shape. Node Expressions too.
HolgerK: SPARQL can be done by the next meeting.
nicholascar: Profiling could be finished in a couple of weeks to a month if necessary.
ieben: Time frame for UI is possibly about a month, there are pending issues such as 1083
edmond: The goal is for the document to be done by late October. We are still on track. A few weeks to be done with the document itself, and then more time to test.
nicholascar: Sounds like we should be able to get the documents written by the end of this month (unclear for Rules).
… We need to get serious about timelines.
HolgerK: Last time in such a WG, it took some time for the step of the process regarding external comments.
nicholascar: We are slightly delayed, Some of the documents have gone through wide reviews, but we need some first implementations before seeking proper external reviews. We might be pushing end of December.
… Hopefully we can get implementations done 3 months before the end of the WG
HolgerK: Do we need weekly meetings again?
nicholascar: for Profiling, probably not.
ieben: Probably don't need weekely meetings with the broader group, just the TF
… for SHACL UI
nicholascar: Call to other TFs to meet as often as necessary to get content done by late August
AndyS: Pre-binding needs to change in SPARQL extensions document to make it non-vendor-specific
AndyS: Rules is ready for wide review.
… We have enough in the document to get started.
… We have longer examples in the document, references and naming and components are being worked on.
… Input requested: optional feature for implementations, which may be moved to the appendix. We probably can't mandate imports.
… More advanced topics: handling of bnodes in templates, document includes a solution.
… Roadblock: naming, has to be done before wide review.
nicholascar: We need to resolve the naming discussion before moving forward.
… Core and SPARQL only have so far gone to wide review.
… Rest needs to be done.
… We have received minimal feedback.
AndyS: Meeting for Rules can be weekly
Rules and naming/shortcode
nicholascar: Suggestions for new shortcode?
… No one has objected to SRL
AndyS: We need to get our suggestion past the RDF WG
… There is an unversioned name and it points to the latest, usually published version now
nicholascar: Any objections to SPARQL-RL?
HolgerK: If it doesn't use SPARQL syntax, I wouldn't call it SPARQL-RL anymore
AndyS: SPARQL-RL is acceptable
nicholascar: No other counterproposals. Let's vote: Does this WG endorse the Rules TF to ask the SPARQL WG and other WGs to approve of SPARQL-RL?
<AndyS> +1
<mgberg> +1
<edmond> +1
<DavidHabgood> +1
<nicholascar> +1
<elianaP> +1
<TallTed> +0
<ajnelson-nist> 0
<YoucTagh> +0
<labra> 0
<ieben> +1
<Livio> 0
<simonstey> +0
<HolgerK> -0.9
nicholascar: There is a clear direction in the votes so we will move forward with this.
… AndyS will ask the SPARQL WG about this.
ajnelson-nist: Given HolgerK's earlier remark about the syntax, we will have to mention the syntax to the SPARQL WG too.
AndyS: SPARQL WG is having weekly meetings right now, by our next meeting the point can likely have been raised there.
simonstey: Why not stick to shapes/shacl RL?
… Is anyone against Shapes/SHACL-RL?
HolgerK: It is still not about shapes, more in general about RDF
AndyS: SHACL rules now have something called shapes rule
… The syntax is not SHACL but RDF. It is only CONSTRUCT
HolgerK: There is a way forward, if the document uses both syntaxes: SPARQL-equivalent would also be allowed
… RDF syntax should probably just be removed
AndyS: The RDF syntax was included by request
nicholascar: Maybe the best thing to do is ask the SPARQL WG
HolgerK: The problem is, the current content of the SPARQL document is also SPARQL rules
AndyS: The full name should be SHACL SPARQL rules
HolgerK: People are going to get mixed up
nicholascar: SHACL 1.2 rules for SPARQL document, the other WG can determine the name for the other rules
… Suggestion is to generalize the naming of that document.
HolgerK: The document has more SHACL features
… SHACL SPARQL rules are the only type of rules that currently exists
… Do we want to put yet another document out there?
HolgerK: We could restart triple rules, they are easy to implement and design
… the other implementation is javascript rules
nicholascar: In other documents, we have tried to include one thing as canonical
… we do need a second rules implementation in that document
HolgerK: The alternative is to not talk about this optional extension point.
AndyS: We have to at least define the type of the extension
HolgerK: Happy to not talk about the example/extension and leave it for a future document
nicholascar: We can ask the other WG what name they find more appropriate
… They probably won't want to publish another document
nicholascar: It's up to the Rule TF to decide
… can we ask the other WG if they are happy with SPARQL-RL or do they prefer another SPARQL/RDF/SRL combination for the document?
simonstey: Maybe Data Shapes Rules
AndyS: the term is being reused already by Holger
nicholascar: Our previous vote stands: SPARQL-RL, or do they suggest something else?
<simonstey> https://
AndyS: there is a website: https://
Ongjen: Could write a short document summarising and sending it around
nicholascar: probably too late for that
nicholascar: we can look at raised issues at the next meeting
nicholascar: no other points, next meeting is in 2 wreeks
… Task force meetings this week
… Core, SPARQL and node expressions not meetintg
… Please finish your content!