<simonstey> zaki, ??P3 is me

<hsolbrig> zakim: IPCaller is me

<hsolbrig> zakim IPCaller is me

<hsolbrig> comma, semicolon, space...

<scribe> scribenick: hknublau

Admin

<Arnaud> PROPOSED: Approve minutes of the 14 May Telecon: http://www.w3.org/2015/05/14-shapes-minutes.html

<pfps> These look fine to me.

RESOLUTION: Approve minutes of the 14 May Telecon: http://www.w3.org/2015/05/14-shapes-minutes.html

<Arnaud> PROPOSED: Approve minutes of F2F: http://www.w3.org/2015/05/19-shapes-minutes.html http://www.w3.org/2015/05/20-shapes-minutes.html http://www.w3.org/2015/05/21-shapes-minutes.html

<pfps> The F2F minutes are a bit variable, but the look acceptable.

RESOLUTION: Approve minutes of F2F: http://www.w3.org/2015/05/19-shapes-minutes.html http://www.w3.org/2015/05/20-shapes-minutes.html http://www.w3.org/2015/05/21-shapes-minutes.html

Arnaud: Next meeting possibly with WebEx

Tracking of Actions and Issues

<pfps> Yes, ISSUE-21 appears to be done

<Arnaud> https://www.w3.org/2014/data-shapes/wiki/Requirements#Selection_by_expression

Arnaud: Closing ISSUE-21 and ISSUE-48

<Arnaud> https://www.w3.org/2014/data-shapes/track/actions/26

<Arnaud> https://www.w3.org/2014/data-shapes/track/issues/raised

<Arnaud> PROPOSED: Close ISSUE-50 adopting Peter's proposal as much as possible

+1

<Labra> +1

<pfps> +1

<simonstey> +1

<iovka> +1

<kcoyle> +1

<ericP> +1

RESOLUTION: Close ISSUE-50 adopting Peter's proposal as much as possible

User Stories

Arnaud: We cannot discuss S55 as Arthur isn’t here, nor S54 as Dimitris isn’t here

SHACL spec

Arnaud: I sent around an email with the proposal

… Jose was out during the conversation so wasn’t able to agree

… Any questions or comments

<Arnaud> PROPOSED: Adopt Holger's draft as the basis for the SHACL specification, leveraging Peter's proposal to improve it, and define a compact syntax for it along the lines of ShEx as described in https://lists.w3.org/Archives/Public/public-data-shapes-wg/2015May/0175.html

Labra: Unclear how ShEx features will be covered

… exchanged emails with Holger, trying to convince him of abstract syntax

… will compact syntax be another section or separate document?

… compact syntax should be based on abstract syntax

Arnaud: I really with we could leave aside the question of how many documents we produce

… all will be on recommendation track

… In Holger’s draft everything is just a draft, similarly any other feature can be added (e.g. the ShEx features)

… WG scrutiny decides per feature

<ericP> hknublau: i don't know why i'm being asked. this is for the WG.

<ericP> ... of course we want one normative defn

<ericP> ... if we want to have a non-normative AS, we can add it as a note

<ericP> ... but the compact syntax should be defined in terms of the RDF representation

hsolbrig: Concerned about the wording “along the lines of ShEx” because ShEx is a family.

… Should use “starting with ShExC” instead to clarify it’s about compact syntax

Arnaud: Agreed, ShExC was the intention.

<Zakim> pfps, you wanted to ask why there is a need for an abstract syntax

pfps: If we refer to ShExC then we need a pointer to a document

… I don’t know what ShExC is (exactly).

<ericP> http://www.w3.org/2005/01/yacker/uploads/ShEx2toy?lang=perl&markup=html#productions

<pfps> If there is a desire to have a particular "thing" for ShEx, then we need a pointer to the document

<ericP> http://www.w3.org/2005/01/yacker/uploads/ShEx2?lang=perl&markup=html#productions

hsolbrig: Is syntax sufficient (we are behind on semantics)

Arnaud: Revising proposal

<Arnaud> PROPOSED: As described in https://lists.w3.org/Archives/Public/public-data-shapes-wg/2015May/0175.html, adopt Holger's draft as the basis for the SHACL specification, leveraging Peter's proposal to improve it, and define a compact syntax starting with ShExC as defined in http://www.w3.org/2005/01/yacker/uploads/ShEx2toy?lang=perl&markup=html#productions

Labra: On Abstract Syntax, this doesn’t need to be the same document, but it’s so important that it needs to be included

… my feeling is that Compact Syntax then maybe it’s easier to talk about features

… but other features of ShEx are not covered by this proposal

… relationships covered by Iovka’s document and we should not lose that

pfps: Not able to vote now because of new link to ShExC is new information

ericP: Current version is similar to submission

… (some details that ericP may want to write down)

<michel> is this vote to accept the proposal as is?

<michel> or is it meant to be the basis by which the language will be developed?

pfps: There were some surprises in ShEx presentation last week

<michel> i'm worried that this conversation is not productive.

Arnaud: I would hate not to have a vote today

<ericP> http://www.w3.org/Submission/2014/SUBM-shex-defn-20140602/#productions submission

hsolbrig: The most stable ShEx version is from Eric’s link.

… I have no problem with saying “starting with ShExC” at a given point

pfps: There have been dramatic changes to ShEx over the last few weeks

… I’d be fine with ShEx as generic label

<Arnaud> PROPOSED: As described in https://lists.w3.org/Archives/Public/public-data-shapes-wg/2015May/0175.html, adopt Holger's draft as the basis for the SHACL specification, leveraging Peter's proposal to improve it, and define a compact syntax starting with ShEx

(I think you should say ShExC)

hsolbrig: ShEx entails more than compact syntax, so that would be fine too

<Labra> http://w3c.github.io/data-shapes/semantics/#abstract-syntax

Labra: Proposal should mention Abstract Syntax

Arnaud: This was already rejected last week

Labra: The vote was in different context

Arnaud: Nobody disagrees with compact syntax, and people are open to base this on ShEx

… We only want one governing semantics, SPARQL as much as possible

<scribe> … Unknown how much is missing by SPARQL, undecided how to formalize the bits that are outside of SPARQL

… Alternative: agree to disagree and not move forward at all

… I understand this is a hard compromise for many

… we need to make one step at a time

<Arnaud> PROPOSED: As described in https://lists.w3.org/Archives/Public/public-data-shapes-wg/2015May/0175.html, adopt Holger's draft as the basis for the SHACL specification, leveraging Peter's proposal to improve it, and define a compact syntax starting with ShEx

… I understand this not the perfect proposal for everyone, but this is how WG’s work

<simonstey> +1

+1

<kcoyle> +1

<pfps> +1

<Labra> -0.5

<ericP> -0.5

<hsolbrig> +0

<iovka> -1

<michel> +0

<iovka> I don't want to obstruct, but a number of us, apparently including Peter, think it won't be possible to use SPARQL templates to produce a semantics that meets the ShEx use cases. Let's note this objection and move on.

iovka: During F2F meeting I proposed to give us time to present, but we didn’t have time

… current proposal does not cover “our” use cases

Arnaud: we have a list of use cases and requirements

… as we move forward, these need to be captured, and accepted

… in my opinion, if this is all then I will have to overrule your objection

iovka: I did not mean to obstruct

… I will make objection more precise

Arnaud: If we find that the foundation is flawed because we cannot address all use cases, we can change everything at any time

… It is always WG’s decision to move forward

… Resolutions stay as long as they are not challenged.

RESOLUTION: As described in https://lists.w3.org/Archives/Public/public-data-shapes-wg/2015May/0175.html, adopt Holger's draft as the basis for the SHACL specification, leveraging Peter's proposal to improve it, and define a compact syntax starting with ShEx

… There is a process for formal objections

… objections will be brought to the attention of W3C management

… Your objection seems to be speculative

… We can have other editors beside Holder

(cannot spell my own name)

… Being an editor is quite a lot of work. Please tell if you are interested in being an editor

pfps: Editors did not have front-line responsibility to answer questions

Arnaud: we need to turn this into Editor’s draft

… Who would like to edit Compact Syntax

Labra: How do we proceed with compact syntax

<pfps> I would think that the Compact Syntax would be a separate document

Arnaud: I suggest a separate document with CS

+1

… What would it take for FPWD?

… Holger should reference ISSUEs in tracker

… raise any open issue from document formally

… when raising issues, you should also suggest solution

+q

hknublau: When could be a FPWD?

… Issues often take two weeks to address

Arnaud: Editorial changes don’t need approval, normative changes need approval

… Cannot give a time frame, expect Holger and Peter to suggest time frame

Raised Issues

… Many of these issues seem already handled

<simonstey> http://w3c.github.io/data-shapes/shacl/#ClosedShape

Labra: Closed shapes were in requirements

… not clear how features interact, e.g. cardinality with disjunction and groups of conjunctions

hknublau: I think we need to be more detailed than the general ISSUEs

Arnaud: Many raised ISSUEs seem to be about mismatch between compact syntax and current draft

Labra: Issues are indeed more general

Arnaud: I suggest editors of CS to do first pass, and relate it to SHACL draft, and find out gaps

… is there a gap already anywhere?

Labra: What would happen with multi-occurrance

… Different allowed values in different constraints

pfps: This problem goes back to ShEx, not a problem with something like SPIN.

… just saying that you don’t know is (unhelpful?)

<pfps> It may be that the document is unclear on how conjunction works. That's a problem. However, is this the case? If not, then what is the problem?

ericP: Is is possible to point to spec now, and then ask, e.g. how to handle multi-occurrence

+q

<pfps> It may be that all these issues arise from the overall shape of the document, which in my opinion is not conducive to a crisp provision of the semantics of SHACL.

hknublau: I have so far only worked against the approved requirements, and that should be the process

Arnaud: I find many raised issues too high-level, they should be more precise

… e.g. Conjunction is not possible yet?

… what you say about interactions of conjunctions with other features

… Title of issue is not informative

(Can we at least open ISSUE-51)?

Labra: Still not clear what I need to do

Arnaud: Make your raised issues more specific

Out of time, happy to follow up in emails

… This was a big day, I hope we can all work better together from now on

<Arnaud> trackbot, end meeting