[CSSWG] Minutes and Resolutions 2010-10-13

Summary:

   - Reviewed status of CSS2.1 implementation reports
   - Reviewed status of CSS2.1 issues
   - Discussed TPAC agenda
   - RESOLVED: Transition CSS Color Module Level 3 to PR
     Disposition of comments:
       http://dev.w3.org/csswg/css3-color/issues-lc-2008.html
     Implementation Report:
       http://www.w3.org/Style/CSS/Test/CSS3/Color/20081014/reports/CR-ImpReport.html

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

Present:
   Tab Atkins
   David Baron
   Bert Bos
   Beth Dakin
   Arron Eicholz
   Simon Fraser
   Koji Ishii
   John Jansen
   Chris Lilley
   Peter Linss

<RRSAgent> logging to http://www.w3.org/2010/10/13-CSS-irc
Scribe: Tab Atkins

Administrative
--------------

   plinss: Anything else on the agenda?
   plinss: 2.1 Testsuite.
   arronei: We published RC2, there's been some feedback.
   arronei: Some progress on the updates.  Elika was planning to publish another
            update at the end of this week, I think.
   smfr: I'm ready to submit an impl report for Safari based on the Oct1 version.
         Should I wait, or go ahead and submit?
   arronei: Go ahead and submit.  We'll send out a list of just the updates.
   arronei: There's maybe 20 updates so far.
   smfr: There's a lot of feedback on the list about tests which have problems.
         Who's repsonsibility is that to roll up that feedback?
   arronei: Elika does that.
   arronei: I've been trying to update the MS tests as they get reported with
            problems.
   arronei: The next publish should have all the feedback done up to that point.
   plinss: The harness on the website didn't get updated to RC2, I suppose.
   plinss: We'll wait for the next update to do that.
   plinss: So, Implementation Reports.
   smfr: Is the right thing to send it to the private or public list?
   plinss: Whatever you're comfortable with.
   plinss: While there are some tests that need to be done in HTML or XHTML,
           does it really make sense to do that entire testsuite twice?
   arronei: I have seen very few differences; maybe 2 or 3.
   arronei: So there is negligible benefit for the vendor, but for reports it
            probably doesn't matter for those handful of cases.
   plinss: For our exit criteria, all I really care about is that it passes
           in HTML *or* XHTML.
   arronei: That's probably fine from a pure spec perspective, because passing
            in one proves that the spec can be implemented.
+Sylvain
   plinss: So that's Google and Apple.  What about MS, Moz, Opera?
   arronei: MS already submitted and posted it.
   <smfr> http://blogs.msdn.com/b/ie/archive/2010/10/08/css-2-1-implementation-reports.aspx
   dbaron: I sent a status update last week.
   <dbaron> http://lists.w3.org/Archives/Public/public-css-testsuite/2010Oct/0051.html
   dbaron: I think I'm down to 500 images left to check.  Then I'll be down
           to manual tests, which I'll probably just do in HTML.
   dbaron: I think the big suite of background combinations could probably
           be done by just querying the OM.
   arronei: Yeah, potentially.
   <dbaron> ... and test the interactive behavior without combinatorics.
   <arronei> MS implementaion report: http://www.w3.org/Style/CSS/Test/CSS2.1/20101001/reports/
   plinss: Do we believe that KHTML and Webkit are different enough to
           count as different?
   arronei: I see a lot of differences.
   plinss: I think that Chrome and Safari are sufficiently close that
           they can't count as our two passing impls, but if Webkit
           and KHTML both pass, can it count?
   smfr: I don't think so - if they both pass, there's a good chance it's
         in shared code.
   arronei: But if Webkit passes and KHTML fails, that still counts as a
            pass, right?
   plinss: Yeah, either one can pass and it's okay.  It's just when they
           agree that we'd count them as one.
   plinss: And the results page on the web harness is starting to show
           something like what we need for the exit criteria.
   smfr: Link to the results page?
   <plinss> http://test.csswg.org/harness/
   <plinss> http://test.csswg.org/harness/results?s=CSS21_HTML&o=0
   <plinss> http://test.csswg.org/harness/results?s=CSS21_XHTML&o=0
   smfr: How are you dealing with the different versions of Webkit, etc.?
   plinss: Right now, if I get a pass in any version of Webkit, Moz, etc.
           I count it as a pass.
   plinss: Later I'll mark them some as flaky if there are mixed passes
           and fails, for manual verification.
   ChrisL: How generic is that?  It would be useful for other CSS3 modules.
   plinss: completely generic; it can be used for any thing.
   <smfr> i hope the user-agent parsing is good :)
   <TabAtkins> smfr, if it's not the base UA string is still stored too,
               so we can reparse.
   <ChrisL> cool so we can use this for modules later

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

   TabAtkins: For Issue 101, the mailing list activity just died down
              yesterday, so I'll synthesize all of that soon.

TPAC Agenda
-----------

   plinss: TPAC, we don't have a lot on the agenda yet.
   TabAtkins: I'll have several drafts I want to talk about by then.
   <ChrisL> request for an FX taskforce joint meeting
   <plinss> http://wiki.csswg.org/planning/tpac-2010
   ChrisL: W3C staff is starting to count up people and plan accordingly,
           so anyone who's coming needs to register now.
   Bert: I've been trying to find people involved in IPDF who could come.
         Haven't found anyone yet, but I'm still trying.
   Bert: I guess most important is the Writing Mode, which is already on
         the agenda, but they're also interested in the text module.

CSS3 Color
----------

   <ChrisL> http://dev.w3.org/csswg/css3-color/issues-lc-2008.html
   ChrisL: I updated the Disposition of Comments just before the call.
   ChrisL: We're in better shape.  Several rejects turned to accept, and
           a few new comments we've gottne accepted. I think it's in good
           shape, and we can justify our remaining rejects.
   <ChrisL> http://www.w3.org/Style/CSS/Test/CSS3/Color/20081014/reports/CR-ImpReport.html
   For 6, I think we can justify the reject, because we have no impls.
   ChrisL: I'd like to put in a new report for Opera, because of changed
           results.
   ChrisL: We could go with what we have, but I think it's fairly easy to
           update our results.
   ChrisL: We have two passes for each test, which is the important thing.
   RESOLVED: Request PR for Color level 3.

CSS2.1 Issues (cont)
--------------------

   plinss: Back to 2.1 issues.
   dbaron: For 154, I didn't see an email to respond to those issues about.
           I think there are a bunch of things I don't like about both of them.
   dbaron: In the first one, the pointer to the line box, because that's
           not defined by a single element.
   dbaron: For the second, it seems unusual for "em-boxes" to represent a
           box with width as well; you usually only care about the height.
   Bert: I think we explicitly said we wouldn't define the "content area" -
         it's a suggestion to take the em height.
   arronei: The problem with "content area" is that it is heavily used
            throughout that chapter, so it should probably be defined.
   dbaron: Font metrics don't necessarily define a box with positions that
           are the em height.
   arronei: So does anyone have update suggestions for it?  I'm not 100%
            sure on what we actually want here.
   dbaron: I sort of see this issue as a nice-to-fix, not a blocker.
   arronei: It's just a detail, helping to explain the text that is already
            there.
   Bert: The current text has some variables - "a" and "d" - and perhaps
         those should be used in the diagrams as well.
   arronei: I can look into that.
   Bert: I agree with David - getting the images would be nice, but I don't
         think we should hold up the spec for them.

   plinss: All right, issue 159?
   dbaron: I haven't got a chance to diff this with the last version and
           see if there are any new comments, but I think that most of my
           comments have been addressed.
   arronei: We've started reviewing it, but haven't finished.
   plinss: We'll give it a little more time for review, then.

   plinss: Issue 199.
   TabAtkins: I gave a proposal for this, but I'm bad at updating the wiki.
              I'll go find it and make sure it's fine today.

   arronei: Bert, can we get an update on your edits?
   Bert: I think there's about two dozen left to edit.  I seem to be doing
         about 6-8 per week, so I should hopefully finish before TPAC.
   Bert: I'm not sure if the edits I've alreayd made are correct yet, but
         assuming everything is fine, I should probably be done by TPAC.

Received on Tuesday, 19 October 2010 21:07:00 UTC