W3C Technical Architecture Group Telcon - 12 February 2007

12 Feb 2007


See also: IRC log


Dan Connolly, Rhys Lewis, Noah Mendelsohn, David Orchard, T.V. Raman, Norm Walsh, Vincent Quint
Tim Berners-Lee, Henry Thompson, Stuart Williams
Dan Connolly
Noah Mendelsohn


Convene, review minutes and agenda

<DanC> minutes 5 Feb

DC: Considering minutes from 5 Feb : http://www.w3.org/2001/tag/2007/02/05-tagmem-minutes

RESOLUTION: Minutes of 5 February 2007 at http://www.w3.org/2001/tag/2007/02/05-tagmem-minutes are approved

DC: Next meeting scheduled for 19 Feb

TVR: US Holiday

Several: at risk

DC: Cancelling 19 February

DO: Regrets for 26 February

NW: Will scribe 26 February

RESOLUTION: Next TAG teleconference will be on 26 February 2007

Issue RDFinXHTML-35

DC: Mail from Harry Halpin at http://lists.w3.org/Archives/Public/www-archive/2007Feb/0015
... Harry is chair of the GRDDL WG
... He's saying we might want to review

From the email:

In order to prevent a "surprise" Last Call, I'd like for the TAG to know that we are going to go, barring any final comments or problems, to request move to Last Call on or shortly after Feb 15th for the following three documents:

1) GRDDL Specification [2]

2) GRDDL Primer [3]

3) GRDDL Use Cases [4]

DC: We were going to last call in Nov, then Jan, etc., but we're really, really doing it soon now.
... We're getting detailed feedback from testers.
... RDFa has a test suite harness moving forward.
... Talking about last call in June, possibly getting Flickr involved.
... Some concern about deploying RDFa in its present state.
... It's cool iff XHTML points to it. Sort of like XML Base.

NW: Right, XML Base works that way. Other specs have to point to it.

DC: So, the story adds up if the XHTML spec is updated.

TVR: Is this true for anything that uses modularization?
... Unless your modules are declared as a profile, you don't have a "follow your nose" path through the specs.
... This was done as an XHTML2 module.

DC: Well, the point is it >needs< to be done as an XHTML module.
... Going back to TAG discussions. Seems we talked about modularization at a meeting (in Cambridge)
... Substitution groups were mentioned as a W3C XML Schema mechanism that supports such modularity.

<DanC> (hot-off-the-presss minutes from the RDFa task force http://www.w3.org/2007/02/12-htmltf-minutes.html )

NM: Or schema wildcards (skip or lax). That's another way of avoiding modifying the container schema.

DC: I hear there's work in progress on XHTML modulariztion.

<noah_> Note that Schema 1.1 is likely to allow elements to be in multiple substitution groups.

TVR: I believe that attempts were made to use schema for XHMTL. I think they had trouble.

NM: I never said schema would or wouldn't meet the XHTML need. I said that in addition to substitution groups, wildcards are a means of writing container schemas that need not be changed when new schemas are written for contained content.

DC: Yes, interesting question. Seems useful in principle of both sides would try harder to make schema work for XHTML.
... Rhys, have you been following this?

RL: We at Volantis did successfully use substitution groups to build a schema for a particular XHTML 2.0 subset.
... Didn't do it myself, but pretty sure it was done.

DO: Seems to be related to the XML section of our versioning finding.
... Maybe we should be sure to describe experiences of XHTML. Maybe will talk to Rhys offline.

DC: Maybe we should ask Mimasa to describe his experiences?

TVR: Also Mark Birbeck.

DC: Suggest I write Mimasa and Mark Birbeck cc: www-tag

NM: Well, there is the public versioning list. Not sure it's still active.

DO: General stuff has been on TAG, XML stuff on schema. Public-xml-versioning is a good choice.

<Norm> http://lists.w3.org/Archives/Public/public-xml-versioning/

RL: No strong view on which list to use.

DO: Rhys, I suggest you look at part 2 of the versioning finding.

NM: Also suggest the draft Dave is doing of a Primer for the schema wg

RL: Dial has some issues related to this.

<Rhys> DI has some of these issues associated with the DIAL spec http://www.w3.org/TR/dial

<DanC> ACTION: DanC to ask Mimasa and Mark Birbeck about feasability of using substitution groups in XHTML modularization, cc public-xml-versioning [recorded in http://www.w3.org/2007/02/12-tagmem-irc]

NM: Suggest one time email to www-tag saying: discussion starting over there.

DO: yes

DC: nah, meeting minutes are enough. Anyone who wants can send www-tag.

<dorchard> Latest version of Schema 1.1 guide to versioning: http://www.w3.org/XML/Group/2004/06/xmlschema-guide2versioning/guide2versioningusingxsd11

Issue utf7Encoding-55

DC: Most recent action was to Vincent to list and describe the issue.

VQ: I sent an email to www-tag right after the telcon. Haven't seen any reaction to it.

<DanC> New TAG issue: utf7Encoding-55 Vincent Quint (Thursday, 4 January)

Vincent's email is at: http://lists.w3.org/Archives/Public/www-tag/2007Jan/0006.html

NM: There was an email exchange with Roy. Not recalling whether the thread reached closure.

<DanC> ban the use and implementation of UTF-7 Roy T. Fielding (Thursday, 14 December)

NW: Issue was raised by Roy.

DC: Don't think we've heard from Roy since.
... If someone who asked us to add this doesn't step up, I'm tempted to drop this.

DO: Interested, but too much to do right now.

VW: There are pointers from the issues list: http://www.w3.org/2001/tag/issues.html#utf7Encoding-55

DC: For now, it's in the "someday" pile.

Issue IRIEverywhere-27 and xmlChunk-44

DC: I had an action to ask Tim if he was satisfied. Sent mail asking: "Are you satisfied?" Suspect he hasn't yet gotten time to read it.
... my action continues
... xmlChunk?

NW: Came up when XSL and XQuery were trying to define deep-equality of XML trees. Those are now at Recommendation, so there is at least one normative definition of deep equality of XML chunks. It may or may not be a useful test in particular cases.

DC: How many tests available?

NW: Not sure.

DC: Wrinkles? Preserves comments?

NW: Lots of wrinkles. Don't know on comments. Not sure I like it.

NM: Defines equality on tree or forest?

NW: Not sure.

NM: Nothing weird like starting in the middle of a tag and ending in the middle of another?

<DanC> (checking http://www.w3.org/2006/xpath-functions# ...)

NW: Nope, not weird like that.

DC: xmlChunk-44 is an open TAG issue?

NW: Yes. We sort of stalled. Lacking some impetus to move forward.

<DanC> action done http://www.w3.org/2004/05/14-tag-summary.html#xmlChunk-44

DC: Norm had action from 2004.

NW: I did something.

<DanC> what he did: http://lists.w3.org/Archives/Public/www-tag/2004Jun/0026.html

DC: Norm's action from 8 May 2004 to propose deep equal function on two XML trees is DONE
... Anyone propose to close the issue?

NW: Yes, sure. There is now an answer. I think we have the impression there is no one right answer.

DC: Did we have a draft finding?

NW: Yes.

DC: Is this a motion to approve that?

NW: Hmm, I need to review that.

<scribe> ACTION: Norm to review draft finding on xmlChunk-44 to see whether issue can be closed and finding approved. [recorded in http://www.w3.org/2007/02/12-tagmem-irc]

DC: Any other thoughs on xmlChunk-44 or IRIEverywhere-27?


DC: Propose to adjourn.
... We are adjourned.

Summary of Action Items

[NEW] ACTION: DanC to ask Mimasa and Mark Birbeck about feasability of using substitution groups in XHTML modularization, cc public-xml-versioning [recorded in http://www.w3.org/2007/02/12-tagmem-irc]
[NEW] ACTION: Norm to review draft finding on xmlChunk-44 to see whether issue can be closed and finding approved. [recorded in http://www.w3.org/2007/02/12-tagmem-irc]
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.127 (CVS log)
$Date: 2007/02/26 18:29:31 $