See also: IRC log
<Roland> http://www.w3.org/TR/xhtml-access/
Roland: The Access module went to FPWD
recently
... We have had a request to look at the selectors API
... and would like comments
... any takers?
Steven: I'm OK with looking at it, but I would think that Mark would be a good candidate
Yam: Is it like CSS elements?
Mark: Yes, a way of selecting elements
programmatically
... OK, I'll take a look
<scribe> ACTION: Steven and Mark to review selectors API by 31 Jan 2008 [recorded in http://www.w3.org/2008/01/09-xhtml-minutes.html#action01]
http://www.w3.org/TR/selectors-api/
Steven: Oh, it's a last call
... so it's a last call review
... While we're talking announcements, Yahoo announced that they are using
XForms for their mobile widgets
Roland: It can lead to a discussion about namespaces for us
Roland: Yam, what is the status?
Yam: Before I contacted our technical team, Steven mentioned that there was an implementation of inputmode somewhere?
Steven: Did I? I don't remember
Shane: Yes, it was Openwave, but I haven't been able to contact anyone there
Yam: Openwave have supplied their browser to
[someone]
... Roland: So we need an implementation of inputmode
... But I assume we need two implementations
Steven: Well, since it is an optional feature,
we only need one working implementation, and show it works correctly
... we don't need two implementations
Yam: I have provided the test report of our implementation
Steven: So we need to create a test report that summarises the test results, shows we have achieved our exit criteria, ans request PR transition
<scribe> ACTION: Steven to bring test results of XHTML Basic 1.1 together [recorded in http://www.w3.org/2008/01/09-xhtml-minutes.html#action02]
Roland: What is holding us up with Modularization?
Steven: Well, there is some problem with a zero-length CR period, so I propose we request CR, and then talk about the length of CR during the call. I would like Roland and Shane to be on the call
<scribe> ACTION: Steven to organise CR call for M12N [recorded in http://www.w3.org/2008/01/09-xhtml-minutes.html#action03]
<ShaneM> XHTML 1.1 Second Edition is dependent on it.
Roland: we'll skip this until Rich is able to be here
Shane: The current DTD doesn't allow the RDFa attributes on the html element
Steven: I would strongly support such an addition
Roland: Which DTD?
Steven: The XHTML 1.1+RDFa
Shane: But this change is not in M12N 1.1
... such as class and id on the html element
... and it is now too late to add it, siunce we are past last call
Mark: Do we really need it for RDFa?
Steven: @about on the root element
Shane: Oh wait, the prose says that id should
be on the root
... so I can fix that, it is an error in the implementation
... and by the way, head doesn't take class either
Steven: Aren't they in the same collections?
Shane: Yes, but we don't use them on head and
html
... we use the id attribute directly
... I will update the implementation right now
... and make the changes in XHTML 1.1+RDFa
Roland: So we are happy with this change?
Steven: Yes
Steven: Not enough people to vote on it today
<scribe> ACTION: Steven create a questionnaire to take CURIEs to last call [recorded in http://www.w3.org/2008/01/09-xhtml-minutes.html#action04]
Roland: Make it two weeks for voting
<Roland> http://lists.w3.org/Archives/Public/public-xhtml2/2007Aug/0009
Shane: I suggested moving h1-6 into the legacy
module for XHTML2
... and then people started asking about the legacy module
Steven: It's a sort of hotchpotch
Roland: We should clean it up and treat it seriously for M12N 2
Steven: It might be a set of legacy modules,
not just one
... since someone might want to use h1-6 but not applet
Shane: Then let's not call it 'legacy' module, but give it a name and mark it deprecated
Steven: Good
... Dates for FtF?
Roland: Feb 18-20
Steven: Feb 19 is my birthday, so we'll have to party that evening!
<scribe> ACTION: Steven to create FtF page [recorded in http://www.w3.org/2008/01/09-xhtml-minutes.html#action05]
Shane: I was wrong, that is fine
Steven: Sounds like a good idea
... Mostly harmless
Shane: Not sure how we do it editorially
... since inputmode is a module in Basic 1.1
Steven: It's not in M12N 1.1?
Shane: Of course not
Steven: That would mean we have inputmode in three places
Shane: It's not a problem to reference the module in Basic
Steven: OK, then let's do that
Roland: Is there value to making a separate document?
Steven: We have three languages that uses it now, so there is a vbalue, but it is not hard to reference it from the Basic spec
Shane: Is it in XHTML2
Steven: Yes, as part of XForms
Roland: Which namespace?
Steven: Not yet resolved by us
... Action?
Roland: It is already there
<Roland> Existing action on Shane to add to XHTML 1.1 SE : http://www.w3.org/2007/07/18-xhtml-minutes.html#action01
<Roland> http://lists.w3.org/Archives/Public/www-html/2007Mar/0038.html
Roland: What is the value space of CURIE?
Steven: URI
Shane: Disagree
Mark: If you compared x:a and y:a and x and y were the same mappings, would they be equal or not?
Shane: I think they are not equal
Steven: I think they should be equal
Shane: I don't think they are compact URIs but scoped names
Mark: In sparql for instance they are compact URIs
<markbirbeck> Mark: They *must* be expanded.
Shane: When looking something up you get it from the DOM and look it up as prefix+extension
Mark: I disagree, you must expand it, because for instance ARIA uses the RDF processing
Shane: Today if I use IE 7, a CURIE doesn't get expanded
Mark: That's an implementation issue
... so doesn't affect the answer to the question "what is the value space of
a CURIE"
Shane: If I say "get elements by role name" what should I use?
Mark: It should be the expanded name
... it is not enough to search for "x:a"
Steven: Since you don't know what prefixes
people have used
... if someone produces a script library, they won't know which prefixes you
have used
Mark: You are searching for a full URI
Roland: We are over time
... let's try and resolve it via email
... who will start the thread?
Shane: Everyone seems to disagree with me, so I
think we are done
... the value space is a URI
RESOLUTION: the value space of a CURIE is URI
<scribe> ACTION: Mark to reply to value space of CURIE email [recorded in http://www.w3.org/2008/01/09-xhtml-minutes.html#action06]
Shane: is it IRI or URI?
Steven: URI I think, since I believe that IRI is a lexical dataspace, whose value space is URI
Mark: Looking at the mail, it looks like we have to discuss further
Roland: We should look at the other issues so
we can go to last call
... so we can't decide to go to last call yet
<ShaneM> I think that there might be some value in preserving the expanded prefix separate from the reference in the CURIE value space
<ShaneM> it would make it easier to search for, for example, all elements that have a role in a specific value space
<markbirbeck> Mark: Ah...the original question points to an answer that is somewhere in between Shane's view and Mark/Steven's view. Peter Johnston says that if CURIEs are to be a superset of QNames then the two parts (prefix and suffix) must be preserved.
<ShaneM> as you say Mark, implementation detail. But I can see some value in it
That's what QNames do, but QNames have a tighter syntax than CURIES