SPARQL Working Group Teleconference -- 23 Jun 2009

<trackbot> Date: 23 June 2009

<AxelPolleres> dialing

<SteveH_> probably...

<LeeF> Scribenick: John

<john-l> Simon: I'm the primary IBM representative.

<LeeF> Scribenick: john-l

UNKNOWN_SPEAKER: I've been working in the Rational group, on a new RDF-based platform, for 4 years.
... We rely heavily on SPARQL, and are very interested in standardizing useful new features as a result.
... We are primarily interested in aggregates, and in updates (in the longer term).

<iv_an_ru> Zakim +01212803 is me

<iv_an_ru> (probably)

<AxelPolleres> http://www.w3.org/2009/sparql/wiki/Agenda-2009-06-23

<AxelPolleres> http://www.w3.org/2009/sparql/meeting/2009-06-16

<LeeF> RESOLVED: Accept last week's minutes at http://www.w3.org/2009/sparql/meeting/2009-06-16

<ivan> regrets for next week, will be at a Dagstuhl workshop

LeeF: Kjetil can probably scribe next week.

AxelPolleres: Anything new with our liasons?

ivan: OWL 2 is now in CR.

LeeF: That includes rdf:text?

ivan: Yes.

<bglimm> rdf:text is now rdf:PlainLiteral in OWL 2

<LeeF> thanks, bglimm, i couldn't remember what the new name was :)

Ivan: We now have a chair for the RDB2RDF WG.

<AxelPolleres> s/???/ivan/

<SteveH_> ??? = ivan

<AxelPolleres> http://www.w3.org/2009/sparql/track/actions/open

AxelPolleres: Can we close any of these actions?

LeeF: I need to swap back in to work on action 16.

<iv_an_ru> Sorry, phone troubles again.

<LeeF> trackbot, close ACTION-42

<trackbot> ACTION-42 Ask team contacts whether http://www.w3.org/TR/sparql-features/ is ok closed

<LeeF> ACTION-44: see http://lists.w3.org/Archives/Public/public-rdf-dawg/2009AprJun/0392.html

<trackbot> ACTION-44 Mail NOT EXISTS example. notes added

<LeeF> trackbot, close ACTION-44

<trackbot> ACTION-44 Mail NOT EXISTS example. closed

<SteveH_> I would like to consider action 40 closed!

AxelPolleres: We had two F&R questions: first, the short name for the document.

ivan: We need to ask the domain admin for a green light on the desired short name.

The group discusses whether we need SPARQL versioning in the short name.

<LeeF> http://www.w3.org/2009/sparql/meeting/2009-06-16#resolution_2

<LeeF> ACTION: Ivan to request sparql-features as short name from Thomas R [recorded in http://www.w3.org/2009/06/23-sparql-minutes.html#action01]

<trackbot> Sorry, amibiguous username (more than one match) - Ivan

<trackbot> Try using a different identifier, such as family name or username (eg. ivan, imikhail)

<LeeF> ACTION: Mr. Herman to request sparql-features as short name from Thomas R [recorded in http://www.w3.org/2009/06/23-sparql-minutes.html#action02]

<trackbot> Sorry, couldn't find user - Mr.

And the group confirms that we already resolved to go with "sparql-features" as the short name.

<LeeF> ACTION: Ivan Herman to request sparql-features as short name from Thomas R [recorded in http://www.w3.org/2009/06/23-sparql-minutes.html#action03]

<trackbot> Sorry, amibiguous username (more than one match) - Ivan

<trackbot> Try using a different identifier, such as family name or username (eg. ivan, imikhail)

<LeeF> ACTION: Herman to request sparql-features as short name from Thomas R [recorded in http://www.w3.org/2009/06/23-sparql-minutes.html#action04]

<trackbot> Created ACTION-48 - Request sparql-features as short name from Thomas R [on Ivan Herman - due 2009-06-30].

chimezie: I haven't been able to review the F&R yet.

AxelPolleres: I think the remaining issues with the F&R document are small.

<AxelPolleres> http://lists.w3.org/Archives/Public/public-rdf-dawg/2009AprJun/0400.html

SteveH_: The CONSTRUCT example is not germane and confusing.

<AndyS> We agreed not to use it last week didn't we?

<SteveH_> I thought so

<SteveH_> http://lists.w3.org/Archives/Public/public-rdf-dawg/2009AprJun/0398.html 2.4.3

SteveH_: In section 2.4.3.

LeeF: I also agree with striking that example.

<AxelPolleres> lternatively, analogously to the SELECT example from before, we can use a subquery with project expressions for this query

<iv_an_ru> (oops)

<SteveH_> http://www.w3.org/2009/sparql/docs/features/#Project_expressions_syntax

<SteveH_> suggestion is to remove first CONSTRUCT example

LeeF: I think including it in the document at this point goes too far.

<AxelPolleres> PROPOSED: remove all below "To return an RDF graph..." in section 2.4.3

<AndyS> Remove just example "CONSTRUCT { ?x foaf:name { fn:string-join(?gn, " ", ?sn) } }

<AndyS> "

<iv_an_ru> I don't like syntax of CONSTRUCT { ?x foaf:name { fn:string-join(?gn, " ", ?sn) } } but I'd keep it.

<iv_an_ru> (we have CONSTRUCT { ?x foaf:name ` fn:string-join(?gn, " ", ?sn) ` } , maybe not the best variant too)

AxelPolleres: We need to remove the offending example and tweak the words to match.

<SteveH_> "To return an RDF graph where the first and family names are concatenated to a full name such project expressions could be used"

<SteveH_> suggestion ^

AxelPolleres: Any objections?

<AxelPolleres> PROPOSED: reword to "To return an RDF graph where the first and family names are concatenated to a full name such project expressions could be used" and remove first COPNSTRUCT example in 2.4.3

<AxelPolleres> RESOLVED: reword to "To return an RDF graph where the first and family names are concatenated to a full name such project expressions could be used" and remove first COPNSTRUCT example in 2.4.3

<AndyS> It is "advice to the editors" isn't it?

<SteveH_> yes

AxelPolleres: We've already taken care of 1 and 2 from the 8 points.

<AxelPolleres> http://esw.w3.org/topic/SPARQL/Extensions/Aggregates

<AndyS> http://www.w3.org/2009/sparql/wiki/Feature:AggregateFunctions#Existing_Implementation.28s.29

AxelPolleres: We should take the list of implementors of aggregate expressions from the ESW wiki.

<AxelPolleres> http://www.w3.org/2009/sparql/wiki/Feature:ProjectExpressions#Existing_Implementation.28s.29

AxelPolleres: Take the existing lists from other pages and paste them into the document.

<AndyS> without the CONSTRUCT case (it's about being concise and focued on select expressions)

Kjetil: What policy should we have for adding implementations? Should we try to be exhaustive?

<AxelPolleres> "The following non-exhaustive list includes some systems addressing this feature"

<LeeF> I share some of Kjetil's concerns

<SteveH> The document will be dated, I don't see the problem

<LeeF> Better to include implementations purely as examples, rather than try to make a comprehensive list

<AndyS> Not exhaustive - but the charter talsk about common extensions and experience which is relevant.

<AndyS> +1 to LeeF

<iv_an_ru> No need to be exaustive, if we're in ;)

<SteveH> yes, not exhastive

<LeeF> what iv_an_ru says is exactly the problem :/

<LeeF> this isn't an Implementation Report

<iv_an_ru> I'd choose two implementations per feature, using "similarity to the spec" as a criterion.

AxelPolleres: Would anyone object to having a non-exhaustive implementation list?

<SteveH> iv_an_ru: there is no spec at this point

KjetilK: I object; I'm concerned about the persistence of the URIs.

<AndyS> Yes

<iv_an_ru> well, what we've specified fo "approximate" syntax and semantics resembles a spec.

<pgearon> I'd like to see the list left

<AxelPolleres> "The following non-exhaustive list includes some systems addressing this feature at the date of publication of the present document"

ivan: The text needs to make it clear that the implementation list could rapidly become out-of-date.
... Also, it needs to emphasize that an up-to-date list *is* maintained on the Wiki.

<SteveH> I don't feel that the document stands on it's own sufficiently if it requires liks to the WG wiki

<SteveH> to justify itsself

<LeeF> Link to which wiki? ESW presumably since it outlives the WG wiki?

<SteveH> +1

<iv_an_ru> +1 for ESW

<SteveH> +1, I also think the docs lifetime is similar to the WGs

AndyS: I think the F&R doc only has a limited span of usefulness, so I don't worry about the deep future so much.
... I would put the links in.

<SimonKJ> +1

<AndyS> We are required to have: "shown to exist in multiple, interoperable implementations" so useful info towards that

LukeWM: Do we still need complex use-cases, with more than one feature, for the F&R?

<LukeWM> http://www.w3.org/2009/sparql/wiki/FRUseCases

<LeeF> Pretty sure it wasn't me requesting them :-D

AxelPolleres: Is there any objection to going to FPWD without the list of complex use-cases?

<LukeWM> sorry LeeF, you're right, it wasn't

<AxelPolleres> PROPOSED: go to FPWD without list of use cases

<pgearon> +1

<SteveH> proposal needs rewording!

<LeeF> indeed

<AxelPolleres> RESOLVED: go to FPWD without list of use cases

<pgearon> Yes, I wasn't seconding that we can go to publication, only that I'm happy not to have the use cases

<SteveH> er?

<LeeF> Intent of the previous resolution is to assert that the group does not wish to wait on complex use cases before publishing FPWD, it is NOT a resolution to publish a document.

<SteveH> nor was it a resolution

AxelPolleres: We discussed the modeling of UNSAID without an actual negation feature in SPARQL 1.0 on the mailing list.

SteveH: I think we can continue with the feature described as is.

<AxelPolleres> 4) in my mail, doesn't need a todo.

AxelPolleres: #5 is just a typo.

SteveH: Adding the TODO to address #6 is fine.

AxelPolleres: I suggest no changes to the doc in response to #7.

SimonKJ: We also support a protocol update, and I'll try to dig out a reference.

<SimonKJ> IBM's Jazz Foundation supports graph update via a RESTful protocol

<SimonKJ> I'll try and find a persistent reference

AxelPolleres: I suggest adding references to these existing implementations to the list for emphasis.
... Is there anything else that would prevent going to FPWD?

<SteveH> +1

<SteveH> ..to ivan

ivan: I think we should make all the above changes and then make a decision.

<LeeF> too many changes to do conditionally, I think

<pgearon> I agree with Ivan

<AndyS> It's FPWD - can be a bit rough. Early is good. "It's agree to publish" not "agree with every detail"

LeeF: Let's make the changes, then at the beginning of the next meeting determine if there is any outstanding hesitation.

AxelPolleres: It'll also be good to have outstanding reviews completed.

<SteveH> my action is still open too, for some reason http://www.w3.org/2009/sparql/track/actions/40

AndyS: Make changes and have reviews?

<LeeF> trackbot, close ACTION-40

<trackbot> ACTION-40 Review F&R document closed

<LeeF> SteveH, i don't know what you're talking about

<LeeF> ;-)

<SteveH> LeeF, the human Zakim :)

Axel arranges for serializing the updates and the reviews.

ivan: I need to finalize the charter for phase 2.
... The charter should include every feature, including the time permitting features. They don't need detailed descriptions, though.

<AndyS> See section 1.1?

<LeeF> http://www.w3.org/2009/sparql/docs/features/#Introduction_features

LeeF: I will take an action to draft text for all of the time permitting features by Thursday.

<AxelPolleres> ACTION: LeeF to draft short descriptions for time-permitting features [recorded in http://www.w3.org/2009/06/23-sparql-minutes.html#action05]

<trackbot> Created ACTION-49 - Draft short descriptions for time-permitting features [on Lee Feigenbaum - due 2009-06-30].

AxelPolleres: LeeF can try to hand off his work to KjetilK for his edits.

<LeeF> Thanks for scribing, John. I'll do the minutes from here.

- DRAFT -

SPARQL Working Group Teleconference

23 Jun 2009

Attendees

Contents

Summary of Action Items

Scribe.perl diagnostic output