[CSSWG] Minutes and Resolutions 2010-09-15

Summary:

   - Peter set up an agenda template for TPAC on the wiki
   - Reviewed status of CSS2.1 issues
   - RESOLVED: Accept fantasai's proposed text for CSS2.1 Issue 173 with option B
   - RESOLVED: Accept proposal for CSS2.1 Issue 203
   - RESOLVED: Mark min/max as at-risk, and mark % in min/max as at-risk. (css3-values)
   - Reviewed Tab's proposal for counter style definitions
       http://lists.w3.org/Archives/Public/www-style/2010Sep/0235.html
   - Discussed how to get implementation reports for CSS2.1 test suite

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

Present:
   Tab Atkins
   David Baron
   John Daggett
   Arron Eicholz
   Elika Etemad
   Simon Fraser
   Sylvain Galineau
   Daniel Glazman
   Brad Kemper
   Håkon Wium Lie
   Peter Linss
   Steve Zilles

<RRSAgent> logging to http://www.w3.org/2010/09/15-CSS-irc
Scribe: TabAtkins

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

   glazou: We need a wiki page for agenda items during TPAC.
   glazou: Elika, can you set up a wiki page?
   plinss: Already done.
   <plinss> http://wiki.csswg.org/planning/tpac-2010

   glazou: CSS 2.1 TestSuite.  Deadline was Sep 15, where are we?
   fantasai: I'm planning to build the testsuite and publish it later today.
   fantasai: We'll call this one RC1.

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

   glazou: Issue 101, Tab sent an email that he wasn't ready.
   TabAtkins: I'll have it before next call.  I can do it by Friday.
   glazou: Issue 154 is on Arron and John.

   <glazou> http://wiki.csswg.org/spec/css2.1#issue-154
   arronei: 154 I've submitted a few images.  I think they're correct,
            though Elika's called out a few things.  I'll see today if
            they need any altering.
   arronei: We may not define all the terms that I use, but we use them
            extensively in the spec.
   arronei: We should probably define them.
   glazou: Action on everyone - review the images for 154.

   glazou: Next issue is 173, on Elika.
   <glazou> http://wiki.csswg.org/spec/css2.1#issue-173
   fantasai: There's a proposal on the list that I sent last night.
   glazou: I think the first part is okay.  We need to decide on the second part.
   <dbaron> what's the url?
   http://lists.w3.org/Archives/Public/www-style/2010Sep/0440.html
   glazou: [lists the options for the second part from the email]
   fantasai: I don't think either choice will end up doing much of anything
             anyway; if anyone's putting a linebreak in generated content
             they need to escape them anyway, and the spec uses \A all
             over the place for it.
   fantasai: So I don't think this is a compat or author issue, so we should
             just choose whichever option is easier for implementors.
   fantasai: I don't have a favored opinion.
   dbaron: I think we handle our generated content the same as DOM text,
           generally.
   arronei: Basically the same in IE.
   <dbaron> we could send it through a different path if we had to first,
            but preferable not to.
   glazou: So option B should match the current practice?
   fantasai: This does mean that it's impossible to represent a carriage
             return in a CSS document.
   TabAtkins: I doubt that's important for anyone ever, so it's okay.

   <glazou> http://wiki.csswg.org/spec/css2.1#issue-197
   RESOLVED: For issue 173, accept part 1, accept option b for part 2.

   <glazou> http://wiki.csswg.org/spec/css2.1#issue-159
   TabAtkins: I haven't had time to process the changes in this draft.
   glazou: I don't want to hold this forever, so please have review ready
           for next week.
   <glazou> http://lists.w3.org/Archives/Public/www-style/2010Sep/0191.html
   TabAtkins: There was discussion during the last telcon about the issue
              that Boris raised with the definition, where a runin is
              clearing one way and the block it runs into is clearing
              another way.
   fantasai: I posted a testcase to IRC and Arron looked at it.
   glazou: So what needs to be done?

   <glazou> http://wiki.csswg.org/spec/css2.1#issue-198
   TabAtkins: We need a new proposal to address Boris' issue.  I can do
              that; we have all the info we need for it.
   ACTION Tab: Post updated proposal to the list.
   <trackbot> Created ACTION-266

   <glazou> http://wiki.csswg.org/spec/css2.1#issue-199
   http://lists.w3.org/Archives/Public/www-style/2010Sep/0433.html
   Tab: Proposal I sent yesterday.  I don't think it actually addresses
        the issue properly, though, so there are some options in the
        email for better solutions.
   TabAtkins: I just need someone to sanitycheck this for me.
   ACTION dbaron: Sanity-check the issue 199 proposal.
   <trackbot> Created ACTION-267

   <glazou> http://wiki.csswg.org/spec/css2.1#issue-198
   <fantasai> http://software.hixie.ch/utilities/js/live-dom-viewer/saved/635
   <fantasai> arronei - what were the test results for that?

   fantasai: Steve, you said during the meeting that there was a problem,
             but you couldn't remember what it was.

   <glazou> http://wiki.csswg.org/spec/css2.1#issue-203
   TabAtkins: This needs layout people to look at this.
   TabAtkins: But elika, arronei, and arronei's coworker all looked at
              this together and thought it was probably good.
   dbaron: I believe the change is fine with me, too.
   RESOLVED: Accept fantasai's change for issue 203.

CSS3 Values: min() and max() at risk
------------------------------------

   <glazou> http://lists.w3.org/Archives/Public/www-style/2010Sep/0233.html
   glazou: Agenda item 3, what should we do with min()/max()?
   dbaron: All of the issues I found with min/max are issues with mixing
           % and lengths.
   dbaron: There were two ways to fix them - one was to remove min/max
           entirely, and the other was to not allow % in min.
   dbaron: There are cases where we reverse percentages for intrinsic widths.
   dbaron: So if you have "width:100px, margin-right:50%", its container
           is 200px to satisfy the conditions.
   dbaron: There's no sensical way to do that with min/max, because there
           are likely multiple solutions and it's difficult to distinguish
           between them.
   dbaron: There were other issues that I'd have to dig up.
   TabAtkins: I say we do what fantasai suggests and mark it as at-risk.
   szilles: That means we'd have to drop the feature, though.  We wouldn't
            have the choice to just drop % from the feature.
   howcome: We can just list both.
   howcome: Also you wanted to move this to Last Call, right?
   dbaron: Yeah.
   howcome: And you're willing to be co-editor?
   dbaron: Sure.
   howcome: I think the two of us should review it through.
   fantasai: Agreed.
   RESOLVED: Mark min/max as at-risk, and mark % in min/max as at-risk.


CSS3 Lists
----------
   <glazou> http://lists.w3.org/Archives/Public/www-style/2010Sep/0235.html
   glazou: Topic, new list-style-types.
   <glazou> +1 !!!
   TabAtkins: I've taken over the list module.  The module extends the number
              of style types from 6 or 7 to around 100, but it still ends up
              still missing a number of minority languages.
   <glazou> http://www.w3.org/mid/AANLkTimg9ULp9vS212+G4f5-Oav63p=Eu+EadAzQmSyk@mail.gmail.com
   <dbaron> http://lists.w3.org/Archives/Public/www-style/2010Sep/0381.html
   TabAtkins: Rather than just forever extending the module with new types
              and still missing things, I've proposed that we can address
              all but 1 of the style-types with a set of 7 algorithms,
              which we can then expose to the user so they can create their
              own list-styles.
   dbaron: I don't think you can get Hebrew simply like that.
   TabAtkins: Actually, it's fairly simple to do.  I have it written down already.
   glazou: Including the forbidden words and such?
   TabAtkins: Yes to the 16/17 thing.  The other forbidden words aren't
              forbidden in a list context, according to some hebrew i18n
              people at google.

CSS2.1 Implementation Reports
-----------------------------

   sylvaing: Is everyone okay with the 10/15 deadline for implementation
             reports?
   TabAtkins: Google should be cool with that.
   dbaron: Not sure if we'll be able to.  It's an awful lot of work.
   * fantasai has some ideas, let's discuss with roc later this week
   sylvaing: Can you be more specific?
   fantasai: It's a question of resources at Mozilla, I think David said before.
   fantasai: I think that Tab and I can possibly help if we can dig up
             HP's system and ask for volunteers to help with the report,
             because then it's easy for one person to submit a small
             number of results.
   sylvaing: So it sounds like we can get an impl report from Google.
             Maybe from Moz.  What about Opera?
   howcome: I can't commit to a date.
   <bradk> What about Apple?
   <smfr> i imagine chrome's results would be equivalent to apple's (both webkit)
   glazou: I'm a bit puzzled, because this is the last effort of a long line.
           Everything depends on this last effort.
   <bradk> I thought there was some forking between Safari and Chrome
   glazou: All the browser vendors are interested in seeing CSS2.1 become
           a Rec, but it *cannot* happen without this effort.
   <smfr> bradk: very little at the css/layout level
   <arronei> Actually Chrome and Safari have different results. If Apple
            can also submit an implementation report that would be nice.
   glazou: I understand the resource issues, but it's very important we get
           this.  It would be a very bad signal if we are late with CSS2.1
           just because of impl reports.
   <smfr> arronei: probably due to different snapshots of the webkit code
   JohnJansen: At the FtF, Moz said it would be okay to do impl reports 30
               days after test suite completion.
   dbaron: I'm not sure those two discussions were related.
   sylvaing: So should we move the date or what?  It sounds like we had
             agreement on the date, but apparently only 2 browsers are
             going to make it.
   fantasai: I think we should leave the date as the target, and I'll
             talk with Tab after the call to see what we can do.
   dbaron: Also I think that looking at an impl report will provide feedback
           on the test suite.
   sylvaing: Right, but we need impl report first.
   glazou: I'd like to remind everyone that if we shift a lot, we'll be in
           very bad shape in the w3c.
   glazou: We publicly said that the impl reports will be ready on the 15
           of oct.  w3c staff read it.
   glazou: So if we shift, we cannot shift a lot.  We still need to be in
           PR before the end of the year.
   glazou: Listening to what david said, I'm scared if it's still possible.
   sylvaing: At TPAC we shouldn't be worrying about 2.1, we should be taking
             it to PR.  I'm scared that it won't happen now.
   glazou: So let's keep the timing right now, and see what Tab and Elika
           can produce by next week.
   sylvaing: And bring this up first thing next call.

Meeting closed.

<dbaron> Is it OK to test a beta for the implementation report?
<glazou> yes
<glazou> as soon as it's publically available
<glazou> and it is
<fantasai> tabatkins: datafile - http://test.csswg.org/suites/css2.1/20100815/testinfo.data
<fantasai> harness info: http://wiki.csswg.org/test/harness

Received on Tuesday, 21 September 2010 09:16:15 UTC