W3C RDF-in-XHTML TF - Telecon 2007-10-18, 15:00 UTC

<mhausenblas> previous 2007-10-11

<msporny> or a list of Zakim commands?

<Ralph> Zakim irc bot instructions

<msporny> I can re-connect if needed, Ralph?

<benadida> next telecon is Friday 26 october, 1400 UTC

<mhausenblas> ScribeNick: mhausenblas

Ben: There are certain edge cases we need to iron out, but we are on a good way

Action Items

-> http://www.w3.org/2007/10/11-rdfa-minutes.html#ActionSummary

Ralph: Syntax Document - process issues are clear for Shane/Steven?

Shane: Steven is currently on holiday - rest offline

Ben: So this will be the draft we propose for publication

<scribe> ACTION: [NEW] All look at tests 46 - 53 and write what you believe the correct triples are [recorded in http://www.w3.org/2007/10/11-rdfa-minutes.html#action10] [DONE]

-> http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2007Oct/0175.html Ben's init the thread

<scribe> ACTION: [NEW] Ben to set up a proper scribe schedule [recorded in http://www.w3.org/2007/10/11-rdfa-minutes.html#action01] [CONTINUES]

<scribe> ACTION: [NEW] Michael to find a more appropriate predicate than foaf:knows for TC46-53 [recorded in http://www.w3.org/2007/10/11-rdfa-minutes.html#action09] [DONE]

-> http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2007Oct/0181.html Manu implicitly resovled it

-> http://xmlns.com/foaf/0.1/#term_maker foaf:maker

<scribe> ACTION: [PENDING] Ben to add status of various implementations on rdfa.info [recorded in http://www.w3.org/2007/10/04-rdfa-minutes.html#action06] [CONTINUES]

<scribe> ACTION: [PENDING] Michael make sure to confirm a design for checking that the ASK SPARQL queries evaluate (yes/no) [recorded in http://www.w3.org/2007/09/06-rdfa-minutes.html#action07] [DONE]

-> http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2007Oct/0171.html Michael's proposal

<scribe> ACTION: [PENDING] Michael to create "Microformats done right -- unambiguous taxonomies via RDF" on the wiki [recorded in http://www.w3.org/2007/08/23-rdfa-minutes.html#action06] [CONTINUES]

Test Cases Review and Approval (46-53, 54-56)

-> http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2007Oct/0170.html Manu's proposed TC validator

<Ralph> scribenick: ralph

Michael: Manu proposed a wonderful implementation of a test case validator
... I'd accept a formal action shared with Manu to implement this
... both as an on-line service and as a downloadable tool
... downloadable version would run locally but still fetch test resources from w3.org

Manu: Ivan's help is necessary for pyrdfa bits

<scribe> ACTION: Michael and Manu investigate with Ivan the implementation of the test case validator proposal on w3.org [recorded in http://www.w3.org/2007/10/18-rdfa-minutes.html#action07]

<mhausenblas> TC

-- test 46

Michael: I've not yet made the foaf:knows -> foaf:maker change; please assume that correction has been made

Manu, Ben: 46 ok

Mark: we're still going to have the @instanceof discussion, right?

Ben: I'm pretty sure Mark and I agree in the case of test 46

Mark: the @instanceof discussion that Ben and I are having does apply to test 46 as well
... I've sent Ben a long email
... I would apply @instanceof to a subject that sits before the @rel

Ben: so you'd have @instanceof inheriting something?

Mark: no, @instance of would create a bnode on the div and @rel and @instanceof both apply to this bnode

Ben: so there would be 2 bnodes?

Mark: yes
... I'm not opposed to changing if there are use cases that require Ben's interpretation

Manu: this will affect all the test cases

Michael: shall we put test 46 on hold pending decision on how @instanceof works?

Ben: yes

-- test 47

Mark: if the question is whether these test are consistent, then, yes; but if the question is whether we agree on the triples, then no
... I think @instanceof should apply to the subject

Ben: I'm confused because I thought these were the ones we agreed on

Mark: if @resource was on its own, then I'd agree that @instanceof applies to it
... e.g. if @rel were omitted from test 47, then we might have the same interpretation

Ralph: are these differences in interpretation of the language in the document or differences revealed by having some actual triples to look at?

Mark: to be fair, I've agreed to let the document proceed even though I disagree with some of it
... looks to me like test 48 works the way I expect

-> http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2007Oct/0175.html "Evaluation of Test Cases 46-53" [Ben 2007-10-17]

Manu: perhaps there is agreement on test 49?

Mark, Ben: test 49 is OK

Mark: I say @about always applies

Ben: this is the one wierd case where I could be convinced to create a bnode
... I agree this test gives the right triples

RESOLUTION: test 49 approved

Manu: perhaps we agree on 50 also?

<mhausenblas> +1

<msporny> ASK WHERE {

<msporny> _:a <http://xmlns.com/foaf/0.1/name> "John Doe" .

<msporny> <http://xmlns.com/foaf/0.1/Person> .

<msporny> }

Manu: the sparql is wrong; should be as posted above

Mark: test 50 is fine with me
... bnode created on the DIV and the property applies to that bnode

RESOLUTION: test 50 approved

<Zakim> Ralph, you wanted to ask about semantics in 49

Ralph: TAG would object to us suggesting that example.org's home page has rdf:type foaf:Person

Michael: add #me then

Ralph: OK

-- test 51

Mark: on test 51 I'd still create the bnode first and attach the property to it

Ben: we've said for a long time that @instanceof is syntactic sugar for a child element

Mark: yes, but not any element
... we'd said that LINK and META apply to the parent

<mhausenblas> we could at least resolve the semantic issue in TC51

<mhausenblas> as Manu put out in his review

Shane: child of LINK or when LINK is a child? LINK's content model is empty

Mark: we used to have @role be a long-handed way of adding predicates to a parent element

Manu: the question is about precedence
... it seems consistent that @instanceof without @about always creates a bnode

Ben: the question is "on what element"?
... I think the property picks up a subject before the bnode is identified

<markbirbeck> s/we used to have @role/we used to have <link rel="xh:role" href="..." \/>/

Ralph: if you need to make assumptions about the order of processing of attributes, does that affect either of these interpretations?

Ben, Mark: we think order of processing does affect both interpretations

Manu: @@scribe didn't capture

Ben: one interpretation makes certain use cases very difficult to write

Manu: but not impossible
... the goal is to make something that is consistent

Ben: no, the goal is to meet our use cases
... how do you write "I know another person"?
... without chaining, this becomes very difficult
... came up first when we considered bibtex
... wanting to give an rdf:type to the author of a paper

Manu: we may have conflicting goals; consistency vs. use cases

Ben: I think we've always had those conflicts

Ben: I agree that Mark's interpretation is definitely consistent

-- test 52

<markbirbeck> I see Ben's issue. I've realised that we are missing one element of the 'old' chaining that we don't have now.

<markbirbeck> In this:

Ben: if I heard Mark correctly, we might agree on 52

<markbirbeck> ...for ever...

<markbirbeck> We used to make the object/subject line up.

Ben: Mark's example in irc is a good example

<markbirbeck> If we were to explicitly 'align' them, would that resolve it for you, Ben?

Mark: in the case of 52, I see @resource working just like @about

Michael: needs #me again

Mark, Ben: agree on 52

Mark: the example I just posted in irc ...
... when we first had chaining (before we'd removed and readded it), it was more explicit that the object of one became the subject of the other
... with one more rule, I think this addresses Ben's use case

RESOLUTION: test 52 approved, pending addition of #me

-- test 53

Ben: one triple is missing from the SPARQL
... add #me again

Michael: in the future, I will clean proposed test cases for semantic bugs first before adding them to the test suite
... for discussion

Manu: consider completely different properties; e.g. type foaf:Document

<msporny> ASK WHERE {

<msporny> <http://xmlns.com/foaf/0.1/Person> .

<msporny> "John Doe".

<msporny> }

Ben, Mark: agree with Manu's triples in irc

RESOLVED, test 53 accepted pending addition of #me

-- test 54

(multiple properites)

Ben, Mark: look good to us

RESOLVED, test 54 accepted

-- test 55

Ben: 55 uses @rel instead of @property

Ben, Mark: 55 looks good to us

RESOLVED, test 55 accepted

-- test 56

Michael: 56 is a smorgasbord

Ben: if it's a reasonable and correct example, I see no reason to refuse it as a test

Michael: I want to understand if it introduces anything new
... it's a matter of time
... I don't see anything new in test 56

Manu: might catch if a developer has inserted a whole bunch of hacks

Mark: consider numbering such tests starting with a higher number

Ralph: how about 200 ? :)

Manu: like torture tests?

Michael: I'll pick 1000

<msporny> ASK WHERE {

<msporny> <http://www.w3.org/2002/12/cal/icaltzd#Vevent> .

<msporny> <http://www.w3.org/2002/12/cal/icaltzd#summary> "Weekend off in Iona" .

<msporny> "2006-10-21"^^<http://www.w3.org/2001/XMLSchema#date> .

<msporny> "2006-10-23"^^<http://www.w3.org/2001/XMLSchema#date> .

<msporny> <http://www.w3.org/2002/12/cal/icaltzd#url> <http://freetime.example.org/> .

<msporny> <http://www.w3.org/2002/12/cal/icaltzd#location> "Iona, UK" .

<msporny> }

<msporny> "rdf:type" needed to be expanded, also missing a "." at the end of

<msporny> the next-to-last statement.

Manu: there are two errors in the SPARQL; missing rdf:type

Mark, Ben: 56 looks good to us

RESOLUTION: test 56 accepted, renamed to 1001

<mhausenblas> Ivan's concerns on approved TC

Michael: Ivan asked us to reconsider tests 11 and 29
... both concern the handling of whitespace

Ben: Ivan has a good point particularly w.r.t. handling of PRE

<scribe> ACTION: Ben enter Ivan's concerns in http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2007Oct/0173.html as issues in tracker [recorded in http://www.w3.org/2007/10/18-rdfa-minutes.html#action08]

Shane: the syntax draft asks "what is the default value for @instanceof"
... this is a schema implementation issue
... do we know an answer?

Ben: can the default be null?

Mark: but that would make every DIV be a bnode

Ben: that's why I proposed null
... since @instanceof changes the way subject interpretation occurs, it should not have a default value
... @instanceof="" should not be the same as omitting @instanceof

Shane: so the answer is there is no default value from a DTD implementation perspective, but there is a default interpretation from an RDFa implementation perspective

<scribe> scribenick: mhausenblas

State of Documents

Shane: Work on Syntax and will forward it to Ralph

Ben: Regarding Primer I worked in a lot of comments, incl. BobDC
... won't be around on upcoming SWD telecon
... Need to request it via mail

Ralph: Also remind the chairs for critical issues

Schedule for next Telecons

Next Friday, 26 Oct 14:00UTC

W3C RDF-in-XHTML TF - Telecon

18 Oct 2007

Attendees

Contents

Action Items

Test Cases Review and Approval (46-53, 54-56)

State of Documents

Schedule for next Telecons

Summary of Action Items