[CSSWG] Minutes and Resolutions 2009-10-21


   - Reviewed last call comments from PF WG
     RESOLVED: that their first comment is out-of-scope for Selectors
     fantasai to respond and answer their other questions.

   - RESOLVED: Remove "Namespaces and down-level clients" (Chapter 11)
               from Selectors

   - Minor Selectors issues: remove :indeterminate tests from test  
     split References section into normative and informative subsections

   - Plan to review state of Selectors implementation reports at TPAC:
     if we have enough implementation reports, we can move to PR  

   - RESOLVED: Yves' comments on the Media Queries and CSS2.1 grammars
               result in no change.

   - Briefly discussed targetting filters issue: further discussion  
     to TPAC and beyond.

====== Full minutes below ======


  CÚsar Acebal
  Tab Atkins
  David Baron
  Bert Bos
  Elika Etemad
  Sylvain Galineau
  Daniel Glazman
  Chris Lilley
  Peter Linss
  David Singer
  Steve Zilles

<RRSAgent> logging to http://www.w3.org/2009/10/21-CSS-irc

Selectors LC Issues

  <fantasai> http://lists.w3.org/Archives/Member/w3c-css-wg/2009OctDec/0035.html
  <fantasai> http://dev.w3.org/csswg/selectors3/issues-lc-2009
  <fantasai> http://lists.w3.org/Archives/Public/www-style/2009May/0065.html
  * fantasai who is minuting?
ScribeNick: fantasai
  <fantasai> dbaron: I don't think this should be done for ARIA
  dbaron: I think ARIA should remain descriptive
  [discussing :checked pseudo-class and ARIA]
  <fantasai> My answer is that it's out-of-scope for Selectors
  <fantasai> Whether a node is checked is defined by the document  
  <fantasai> whether that's defined by the HTML4 spec
  <fantasai> or HTML5+DOM+SVG+ARIA+something-else
  People seem to agree with this

  Chris: We asked them a question about ::before and ::after, they  
don't seem
         to have responded.
  fantasai: so who should respond?
  Daniel: you
  Bert: Make sure they know it is an official response
  ACTION fantasai respond to Janina

  <fantasai> http://lists.w3.org/Archives/Public/www-style/2009Apr/0006.html
  fantasai: This was raised against implementors, actually, but filed  
  fantasai: I think it's out-of-scope for Selectors, just wanted to  
check that
            that's acceptable
  Tab: Seems fine to me
  dbaron: I'm not sure how reasonable Anne's suggestion will be for  
our UA
          style sheet, but that's another story
  Peter: No objections to treating as out-of-scope

  Issue 1.
  Summary: Remove suggestion to use \: to match namespaced XML in down- 
level UAs
  From: Krzysztof Maczynski
  Comment: http://lists.w3.org/Archives/Public/www-style/2009Mar/0098.html
  Response: http://lists.w3.org/Archives/Public/www-style/2009Oct/0239.html
  Comment: http://lists.w3.org/Archives/Public/www-style/2009Oct/0251.html
  fantasai: Anne suggested removing the entire section. I don't mind
  Peter: I agree as well, and I original wrote the section
  Peter: It was mainly an argument to Microsoft 10 years ago  
explaining why
         escaped colons was a bad way to handle it and that we needed a
         Namespaces spec
  ACTION fantasai remove Chapter 11

  fantasai: That leaves 3 more issues, all on defining things defined in
            2.1 but not here
  fantasai: I don't have any concrete proposals here
  Bert: Point to CSS2.1 and HTML and XML specs
  fantasai: For ::first-line (issue 10), I can split out a lot of the
            CSS-specific stuff, but I can't figure out a general way to
            explain that we're referring to the first line of a block
  <ChrisL> say the host language defines it
  fantasai: I'll try to fix that and post my edits to www-style for  
  Chris: For Issue 6, refer to XML for the first 3 and to the DOM spec  
for DOM
  Chris: e.g. DOM3 Core

  Chris: I have a question about :indeterminate
  Chris: The spec has a note about :indeterminate, no normative text,  
but we
         have a test in the test suite
  Chris: We should either define :indeterminate, or drop the test and  
  dbaron: Isn't this the result of agreeing to drop it in the past and  
          having changed the test suite to match?
  fantasai: I don't know, I wasn't editing the spec at the time. Daniel?
  Daniel: :indeterminate was not in the original Selectors spec, it was
          added by Tantek
  Chris: I think the tests mostly fail, so I suggest dropping those  
  ACTION fantasai drop :indeterminate from tests

  Chris: Why aren't references split into normative and informative?
  fantasai has no clue
  fantasai: I can make that split
  fantasai: I'll have to reword the PCDATA/etc. to not require normative
            references to HTML/XML
  Steve: Is there a section that says what things are language-specific?
  Chris quotes section 13 - Conformance
  Peter: are all concerns answered?
  Chris: So we want to move to CR, how long should it be in CR/ what are
         status of implementations?
  dbaron: Why can't we move to PR?
  dbaron: I thought we had already met our CR exit criteria
  fantasai: Do we have implementation reports?
  dbaron: I thought we already had implementation reports
  fantasai can't find them
  fantasai: If we don't have implementation reports in the next 2 weeks
            then I think we should move to CR
  Bert has links to the implementation reports
  <dbaron> http://www.w3.org/Style/Group/documents.html
  <Bert> FF 1.5: http://lists.w3.org/Archives/Public/www-archive/2007Jul/att-0046/implementreportFx15-20070725.html
  <Bert> Opera 9.5 http://lists.w3.org/Archives/Public/www-archive/2007Sep/att-0055/selectors-report.html
  <Bert> Konq 3.5 http://lists.w3.org/Archives/Public/www-archive/2007Sep/att-0054/implementreport-konqueror.html
  <Bert> Amaya 9.55 http://www.w3.org/Style/Group/css3-src/css3-selectors/amaya-9.55-report.html
  fantasai: FF1.5 is old.

  <fantasai> http://www.w3.org/Style/CSS/Test/CSS3/Selectors/current/
  <fantasai> which is linked from the test suite
  <fantasai> I haven't dropped the :indeterminate tests yet
  <fantasai> I'll do so soon
  Chris: The implementation report template is awkward, you have to  
         the same test multiple places
  We will discuss the transition at TPAC
  Peter: We'll have implementation reports by then hopefully?
  * fantasai is not taking responsibility for implementation reports
  Chris is working on an Opera one
  dbaron is working on Mozilla (?)
  fantasai: Can we get one for WebKit?

  <dbaron> fantasai, the current implementation report template has one
           link to the 20090603 test suite, but the individual test  
           are all to the 20060307 test suit (note year!)
  <dbaron> fantasai, I'm going to assume that's wrong and search-replace
           the whole thing before running it
  <dbaron> fantasai, I'm going to prepare a revised  

Syntax and Grammar

  Peter: Media Queries syntax from Yves
  <plinss> http://lists.w3.org/Archives/Public/www-style/2009Oct/0095.html
  Bert: First issue was about using functional syntax for media queries,
        not much chance for that
  Bert: Second issue was to include the parentheses in the CSS2.1
        Appendix G grammar
  Bert: Because that's what he uses to parse
  Bert: I'm not in favor. It's not wrong to add it, but it's not  
  fantasai: If Yves needs it for his implementation, then he can add it.
            We don't need it for the spec, so we dont' need to edit  
the spec
  dbaron: The Appendix G grammar is only a description of what's in  
          it's not a guide for implementations
  dbaron: It's sometimes useful where the spec fails to give a grammar
          elsewhere, but we should really try to eliminate those
  Bert describes what Media Queries does to the grammar
  Bert: I'm happy to do nothing here
  RESOLVED: Closed no change

@page and the CSSOM

  <plinss> http://lists.w3.org/Archives/Public/www-style/2009Oct/0094.html
  fantasai: I have no idea, but the @page rule with its nesting  
            can't be changed at this point
  fantasai: Also we have nesting for @media...
  * fantasai notes that Anne is not on the call

Targetting Filters

  Chris: I'm concerned that we would replace the explosion of properties
         with the explosion of pseudo-elements
  Tab: There was some discussion of this with Brad and roc
  Tab: We don't want the combinatorial explosion
  Tab: But other ways of naively doing it break the CSS model for values
  Daniel: Microsoft introduced filters long long ago
  Daniel: It would be cool to provide a replacement for this  
proprietary feature
  Chris: SVG filters is being split out into a separate file, to allow  
         specs to incorporate them
  * sylvaing wishes he could use consistent CSS across HTML and SVG for
             transitions, animations, transforms, filters....
  Daniel: After transitions, animations, and transforms filters will  
be the
          next feature requested by web designers
  Tab: So hash out more of this on the list? I don't think this is  
       we can resolve on the call
  * fantasai agrees
  Tab: It's a big issue. No matter what we do, it will be something  
       and complicated to work out
  Peter: Any benefit to working F2F at TPAC?
  Daniel: If we follow Chris's suggestion, it would be something to  
          with SVG
  fantasai: SVG won't be meeting at TPAC
  Chris: Some members will be present, but not the group
  Chris: Erik would be the best person to talk to, since he's editor for
         Filters spec
  Deferred to TPAC and beyond

Meeting closed.

Received on Friday, 23 October 2009 20:11:51 UTC