XHTML2 Working Group Teleconference -- 21 Jan 2009

<trackbot> Date: 21 January 2009

<scribe> scribe: Steven

CURIEs

Roland: Curies have gone to CR

Steven: Let's talk about test suites

Roland: Does the RDFa test suite cover all features?

Shane: It covers all features, there is a datatype that isn't tested, not a feature per se
... I've grabbed the test harness and I am extractinng the CURIE tests
... Harder than I had expected

<Tina> Apologies; time slipped away and my phone is not behaving :(

Shane: it's doable

Steven: Moving the harness to W3C space?

Shane: No, and they use a different version of PHP
... Ralph has already tried and failed
... when Ralph has fixed it for RDFa, then it will be easy for us

Steven: Good news

Roland: Timescale?

Shane: Not sure yet

Media types

Roland: I see we had comments

Shane: I sent it to Steven for publication

Steven: I sent it on to be published, and it had arrived, but Monday was a holiday, and yesterday sort of too, so maybe it's just backed up

Shane: We also got a comment about @lang, and I replied that we can't add it to existing family members, but XHTML 1.2 could

Roland: it is definite we can't get it back in?

Shane: It would break conformance

Roland: Maybe the Mobile web people could do the grunt work to get it into the system,

Shane: So you think it is OK to break conformance?

Roland: Maybe, not sure.

<mgylling> C.7. The lang and xml:lang Attributes

<mgylling> Use both the lang and xml:lang attributes when specifying the language of an element. The value of the xml:lang attribute takes precedence.

Steven: If we say that @lang has no meaning, it wouldn't actually break anything

Roland: Why don't we test the waters with the mobile people?

<scribe> ACTION: Roland to talk to the mobile people about allowing @lang [recorded in http://www.w3.org/2009/01/21-xhtml-minutes.html#action01]

<trackbot> Created ACTION-45 - Talk to the mobile people about allowing @lang [on Roland Merrick - due 2009-01-28].

Steven: The doc is in the pipeline, as soon as it happens, I'll let you know

Roland: And the other docs are in the pipeline waiting for this to be published

Steven: Yes, that's right

Shane: If someone in autority would say that adding @lang would not vioalte PER, it would be 2 mins to add it to the spec

Roland: We will carry on as we are for now

XHTML2

Roland: Shane has asked some questions
... so which version of XForms?

Steven: 1.1

Shane: And how does that fit in with XMNL Events 2?

Steven: We need to check that they are consistent
... XML Events 2 is based on XForms 1.1, so it should be the same, but we need to check
... XForms 1.1 still references XML Events 1, and adds stuff into its own spec (like RDFa does with CURIEs)

Markus: Is the XForms 1.1 schema chameleon?

Steven: Yes
... so charset and encoding are related (XML Events)

Shane: So you think we should allow @charset

Steven: Yes, for legacy use (mainly for the javascript + @implements" case)

[Discussion on whether charset or encoding is a better choice]

Roland: What is the difference? One is a single value and the other a list?

Steven: Yes
... you should look at the definition of charset in HTML4, it really isn't very useful
... which is why we made @encoding what it is
... to make it actually useful
... I suggest we should say @charset has no meaning
... and only there for legacy use

Roland: It is a compormise I can live with
... so I can specify both if I want
... So the next question is what to do with the script module
... I think there should be one module for XHTML 1.2 and 2.0

Steven: Isn't script in XML Events 2?

Shane: Yes

Roland: But it is a separate module
... caption element

Shane: There are other issues that we need to discuss as well
... but I may have confused people here
... we use caption on object
... so my question is really is it OK to split caption out as a module
... and Mark then pointed out that we have label as well
... and I don't care which we use

Roland: We have caption in older versions of XHTML
... and label has other parameters, like ref
... so an unadorned caption in object would be simpler

<ShaneM> http://htmlwg.mn.aptest.com/htmlwg/xhtml-m12n-2/mod-list.html#edef_list_label

Shane: The label in the list module is a different one (no @ref)

Markus: Perhaps it should be caption on lists then...

Steven: It would certainly be confusing for the user if label had different meanings in different places

Shane: Good point. I think that caption should be drawn out into its own module

Steven: Agree

<mgylling> +1

Shane: So should we use caption in lists?

Markus: What does this mean for img then?

Steven: Are you suggesting a caption child of img?
... from a processing point of virew, img and span are identical

<ShaneM> http://htmlwg.mn.aptest.com/htmlwg/xhtml-m12n-2/mod-text.html#cs_Text

Shane: The question really generalises to "should caption be able to be a child of any element"
... or "if we define the caption element, should it be in the text content set"

Markus: Or the title element, like the title attribute

Steven: The relationship between @title, caption and label is difficult
... it needs cleaning up

Shane: I think caption on image is a fine idea
... I should record it as an issue whether it should be allowed more generally

Steven: HTML5 has a figure element, with a child of <legend>

<ShaneM> ACTION: Shane to define caption module and add issue about whether caption should be included in the Text content set. [recorded in http://www.w3.org/2009/01/21-xhtml-minutes.html#action02]

<trackbot> Created ACTION-46 - Define caption module and add issue about whether caption should be included in the Text content set. [on Shane McCarron - due 2009-01-28].

Shane: We should put Tina's XHTML2 issues at the top of next week's agenda

[ADJOURN]

- DRAFT -

XHTML2 Working Group Teleconference

21 Jan 2009

Attendees

Contents

CURIEs

Media types

XHTML2

Summary of Action Items

Scribe.perl diagnostic output