XHTML WG Call -- 12 Dec 2007

whatever

yeah - but zakim wasn't on the channel yet

<oedipus> ah - i should stick to my own systems' problems

<markbirbeck> Hello all. I am sending my apologies, because if I don't get my changes to the RDFa draft finished in the next couple of hours, I won't be able to do anything with them until late on tomorrow.

<markbirbeck> I'll stay on IM though, in case there is anything I can help with.

<scribe> scribe: ShaneM

<Roland> Agenda: http://lists.w3.org/Archives/Public/public-xhtml2/2007Dec/0009.html

curie syntax

dont have a quorum to resolve to publish it. Roland suggests that we poll via e-mail.

no one on the call has a problem taking it to last call.

RESOLUTION: Those of us on the call are happy.

<scribe> ACTION: Roland will send an email poll out to get us to last call on CURIEs. [recorded in http://www.w3.org/2007/12/12-xhtml-minutes.html#action01]

XHTML basic testing

Yam is not here, but he has sent a report to Steven. We need to get Steven to deal with it.

a role for object - http://lists.w3.org/Archives/Public/public-xhtml2/2007Nov/0015

@role on object element

the role attribute is on everything, so it is already on object.

One reason alessio wants this is to help improve content negotiation in objects.

Rich: are they asking that we extend our collection of XHTML values?
... Its improtant that we remember our existing values are about document sections.
... The values listed here are sort of a random selection. If we are going to look at this we should do a comprehensive stufy.

Gregory: Alessio is really asking if this is something he should investigate further

Rich: is this really something that the XHTML 2 Working Group should be defining

Roland: Not a big fan of object. Seems a bit of a bucket.
... might be useful to learn from the work that HTML 5 people are doing.

Rich: would need to spend a lot more time on this to do it justice.

CURIE Comments

There are no outstanding comments.

Role last call issues

<oedipus> rich: http://htmlwg.mn.aptest.com/cgi-bin/xhtml2-issues/Role?user=guest;statetype=-1;upostype=-1;changetype=-1;restype=-1

There are 10 open issues

<oedipus> http://www.w3.org/2007/11/09-xhtml-minutes.html#item01

<oedipus> RESOLUTION: allow role to be chameleon (http://www.w3.org/2007/11/09-xhtml-minutes.html#item01)

Issue 8017 - what should a user agent do when looking at multiple role values on an element

<oedipus> "allow role to be chameleon" is the only item noted as resolved in f2f minutes

Rich: browser vendors are going to use the first one they recognize for navigation...
... There might be an extension mechanism for this in ARIA in the future. Role is also used in middle ware, so it is not about the user agent all the time.

something like this: The role attribute specification does not require any behavior by user agents. So the interpretation of values by user agents is outside the scope of this specification.

Issue 8019 - do not use CURIEs for role values

scoped values are an extension mechanism, and we need some mechanism for ready extension of roles. If CURIEs don't work out, we will have to consider other mechanisms, and full URIs are worthy of consideration.

<scribe> ACTION: Send formal reply to 8019 [recorded in http://www.w3.org/2007/12/12-xhtml-minutes.html#action02]

ensure the submitter knows that the CURIE spec is separate and will be referenced by role in its next draft.

<oedipus> "I wonder in which namespace that non-qualified role values are interpreted." "we'll have hard time finding which one's the namespace or which one's the Class identifier"

issue 8020: shane will deal with

issue 8021: editorial

issue 8023: al gilman's comments

Issues 1 and 2 are dealt with via chameleon namespacing.

Issue 3 has been closed through negotiation.

<scribe> ACTION: Gregory to respond to issue 8023 formally so we have closure. [recorded in http://www.w3.org/2007/12/12-xhtml-minutes.html#action03]

issue 8024:

There is an issue about CURIEs and how they behave when there are errors. That should be split out into a separate issue.

<scribe> ACTION: Shane to split 8024 into two issues. [recorded in http://www.w3.org/2007/12/12-xhtml-minutes.html#action04]

CURIE spec does not deal with the issue of what happens when CURIEs are broken.

Editorial comments in the issue are valid, and will be addressed.

<oedipus> quote from PF telecon meeting 26 November 2007 (http://lists.w3.org/Archives/Member/w3c-wai-pf/2007OctDec/0224.html): AlGilman: "AG: tell XHTML2 WG not to worry about title as predefined role, PF will do it"

Issue 8025:

comments on the DTD implementation. agreed. will fix.

Issue 8026:

removing CURIE defintiion from role.... sure.

comment 2 - overcome by events

<oedipus> take-away comment: "Given that all prefix expansion is seem aligned with in-scope namespace decls, I'd expect the defaulting to be similarly aligned. Also, given the potentially generic nature of CURIEs as a construct in other languages, I would rather its defaulting was not bound to the the XHTML namespace"

comment 3 - define a schema datatype for URIorCURIE. will do.

roland: we should include the schema implementation; possibly as non-normative for now

<scribe> ACTION: Shane to add a non-normative schema implementation to CURIE spec [recorded in http://www.w3.org/2007/12/12-xhtml-minutes.html#action05]

issue 4 - correct, the definitions of CURIE were inconsistent and the one in rdfa-syntax is the right one. Overcome by events.

issue 5: the default is not in the xhtml vocab space for CURIEs generically, just with regard to the role attribute when used in the XHTML namespace.

issue 8027: agreed to allow chameleon

issue 8028: tag comment about curies. We have said the comment is not in scope because it is about CURIEs, not about how @role uses CURIEs.

- DRAFT -

XHTML WG Call

12 Dec 2007

Attendees

Contents