[CSSWG] Minutes and Resolutions 2010-07-21

Summary:

   - RESOLVED: Republish CSS Snapshot 2007
   - RESOLVED: Publish Media Queries after updating note about media queries
               in HTML to point to HTML5
   - RESOLVED: Change "A minimum of another six months of the CR period
               must elapse" to four weeks for Media Queries
   - Discussed CSS2.1 Issue 110 (abspos in table box construction)
       http://wiki.csswg.org/spec/css2.1#issue-110
   - RESOLVED: Proposal accepted for CSS2.1 Issue 138 with correction in reply
       http://wiki.csswg.org/spec/css2.1#issue-138

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

Present:
   David Baron
   Bert Bos
   Arron Eicholz
   Elika Etemad
   Simon Fraser
   Sylvain Galineau
   Brad Kemper
   Chris Lilley
   Peter Linss
   Steve Zilles

<RRSAgent> logging to http://www.w3.org/2010/07/21-CSS-irc

Scribe: szilles

Republish 2007 Snapshot, MQ CR
------------------------------

   Peter: Proposed to republish 2007 Snapshot with css-style-attr included
   Bert: no objection to re-publishing
   RESOLVED: Publish CSS Snapshot 2007

   Review sections 2 and 3 of MQ
   <ChrisL> In media queries, is it still true that "HTML has not yet been
            normatively updated to use media queries in the media attribute."?
   <ChrisL> thought html5 had been updated to use media queries
   <ChrisL> http://www.w3.org/TR/html5/common-microsyntaxes.html#mq
   Peter: Everythng else in MQ refers to HTML4
   Chris: not much point in refering to HTML 4.01 at this point
   fantasai: There is no reason to have a forward reference to HTML5
   <dbaron> If the reference is explaining the history, it should probably
            refer to the spec that was there at the time.
   Chris: Since the reference is non normative, there is no reason not to
          update the references to HTML5 to be more timely
   dbaron: the references to HTML4 are mostly indicating the history of
           the development of the MQ spec
   Bert: HTML5 is only a draft so we cannot assert that it is done and will not change.
   Chris: we could say that "the current draft of HTML5 has added ..."
   RESOLVED: Publish mediaqueries after updating note about media queries
             in HTML to point to HTMl5
   fantasai: I believe the test suite exists it just needs to be posted
   Bert: Pub Rules require some date
   Chris: four weeks is a reasonable time; it would not constrain us and
          the test cases still need to be posted
   RESOLVED: Change "A minimum of another six months of the CR period
             must elapse" to four weeks
* bradk has to leave the call now. Bye.

CSS2.1 Test Suite
-----------------

   Arron: Sent one mail on the likely invalid tests; need people to look
          at to see if they agree
   Arron: About to send tests that do not pass in any browser; need people
          to check the tests in their area of responsibility
   Arron: Would love to have responses ideally by next phone call, but at
          least by the call in two weeks.
   Peter: Has anyone been working on their implementation reports?
   Arron: is there a template for the report?
   fantasai: no, but I could make one if we agree on what it should have
   Arron and fantasai: to talk offline to agree on the format for the report
   Peter: Is the test harness that was developed still functional?
   fantasai: not at all sure; I have not looked at it.
   Peter: is it worth spending some time to get that running again; it was
          designed to gather statistic and be a start to an implementation
          report
   fantasai to contact Tab for help on this; Peter can help and contact Matt Bonner

CSS2.1 Issues
-------------

   <plinss_> http://wiki.csswg.org/spec/css2.1#issue-110
   dbaron: I believe Boris has objected (threatening formally) to this proposal
   dbaron: The objection is that there is currently interoperable behavior
   dbaron: The behavior is that abs pos'd elements generate cells
   fantasai: I cannot think of any other place where moving an abs pos
             element out of normal flow generates normal flow content
   fantasai: the choices seems to be between defining the behavior the
             way authors would want it or the way that implementers
             would want it
   SteveZ: How common is having a abs pos element as the direct content
           of a table cell or equivalent situations?
   SteveZ: Is this enough of an edge case; is it sufficiently uncommon
           that leaving the existing behavior is OK if not really desirable
   Peter: That is my feeling
   fantasai: There is nothing in the spec that would cause an abs pos
             element to have an effect on surrounding normal flow content
   <fantasai> But all the browser implementations implement abspos by
              introducing a "placeholder" in their formatting structure
   <fantasai> In most cases, this placeholder is not detectable
   <fantasai> but in table box generation it is detectable
   <fantasai> And so bz is saying to codify its existence, at least in
              this one case where it is detectable
   <fantasai> because all implementations chose to implement abspos this
              way and we have interop on it
   <fantasai> even though it makes no sense from the spec's point of view
   Peter: Does anyone have an objection to BZ's proposal
   fantasai: I would like more input from the other implementers as to
             which way they would like to go.
   Chris: Other implementers, like Prince?
   fantasai: yes, and Antenna House
   <Bert> Is it the case that   <tr><td psoition:absolute>a <td>b <td>c
          looks the same as  <tr><td>b <td>c    ?
   <fantasai> In bz's proposal <tr><td>a<td position:absolute>b<td>c</tr>
   <fantasai> looks the same as <tr><td>a<td><td>c</tr>
   <fantasai> Also, <tr><td>a<td position:absolute>b<td position:absolute>c<td>d</tr>
              looks the same as the above,too
   Peter: Should this case in fact be codified or should it just be left
          undefined?
   Chris: That depends on whether the other implemetations are willing
          to change; if not we would have to leave it undefined.
   dbaron: It may be the case that other implementations do not even
           implement this part of the spec.
   SteveZ: we need to send a note to the implementers we are aware of
           and ask which proposals they can live (and would be willing
           to change to)
   ACTION fantasai: Write proposal to deal with abspos in table box
                    generation according to bz's proposal

   <plinss_> http://wiki.csswg.org/spec/css2.1#issue-138
   Issue 138
   Peter: this is awaiting feedback from Alex
   Sylvain: Alex thought the proposal was reasonable
   <plinss_> http://lists.w3.org/Archives/Public/www-style/2010Jul/0329.html
   fantasai: I recall the float should move with rel pos of the containing block
   dbaron: Tab has revised the proposal
   Question about the "contained in the block box" should not that be
     contained in the in-line block
   <dbaron> also http://lists.w3.org/Archives/Public/www-style/2010Jul/0331.html
   dbaron: there are two messages in the thread to that effect
   <dbaron> I'm happy with the revised proposal.
   <smfr> me too
   Bert: it describes what I expect so I am happy to
   RESOLVED: Proposal as modified by the two message is accepted

Meeting adjourns at 10 PDT
<RRSAgent> http://www.w3.org/2010/07/21-CSS-minutes.html

Received on Wednesday, 21 July 2010 20:36:33 UTC