SV_MEETING_TITLE -- 29 Sep 2010

<glazou> rhâââ european bridge saturated, as usual

<jdaggett> thx

<glazou> note to self: extra item http://www.w3.org/mid/4C9CC808.7010402@nokia.com

<Ms2ger> np

<scribe> ScribeNick: TabAtkins_

glazou: Request from Art Barstow for extra agenda item.

<glazou> http://www.w3.org/mid/4C9CC808.7010402@nokia.com

glazou: Now, CSS2.1. Where are we with the test suite, IR, etc.?
... First, test suite.

fantasai: There were a bunch of edits checked in by gsnedders about updating metadata.
... And some other fixes.
... I'm not certain every issue has been addressed. If requested, I could publish a snapshot today.

glazou: What about IR?

smfr: I anticipate that Apple will be able to do test reports.
... I just need ot know which version to test - we probably won't be able to do this more than once.

arronei: If we were to publish an update, we'll list only the differences from the previous, so you could just run today's testsuite and then run the handful of tests from the new suite.

smfr: That would be fine.

glazou: So for now take the 20100917 version, linked form the testsuite page.
... I pinged Opera and Moz to ask if we could have an impl report from them.

howcome: We should be able to do it, I just need to get the right people internally.

glazou: When do Opera and Apple think they could submit it?

howcome: That's a much harder question to answer. ^_^

glazou: I understand. Mid november, that would be fine. Mid Feb, not so fine.

smfr: We anticipate by the end of Oct, possibly by mid Oct.

howcome: I can't give a date at this point.

glazou: Understood; it's not super urgent right now, but we need more IRs more tests that aren't passing in at least 2 yet.

smfr: I implemented a harness that lets you run through the test suite, with a few extra features.

glazou: Great; if you could submit it to the group or www-style it could be useful.
... I would like to warn all members of the WG that some tests require a change of character encoding for the doc, and some browsers reload when they do that.

smfr: Do we have a flag for tests that require a change in character encoding?

TabAtkins_: I don't believe so.

fantasai: Do you mean changing character encoding via UI, via HTTP, via @charset, via <meta>...?

glazou: The ones I was mentioning were about changing it in the UI.

<ChrisL> The ones where usrer interaction is required

fantasai: I think we should mark those as "need user interaction" or do a separate flag.

smfr: I'd prefer a separate flag.

fantasai: Okay.
... Do we need one for general "needs special settings", or should we split it out?

<glazou> really glad

smfr: I think it's okay to share a flag with user styles.

fantasai: Okay. I can't do that tonight, but I'll look into it.

glazou: Enough about the test suite. Bert, you had a list of edits to do.

Bert: I didn't do as much as I had hoped to do. I've done 6, and I think 20 left.
... I assume I can do the remaining edits in 2 weeks.

glazou: Anything blocking any of these?

Bert: I don't think so.
... But I need to review some of Elika's reviews of my edits.

TabAtkins_: I will have the rest of my 2.1 issues on the list by today.

szilles: I sent a message to the list yesterday pointing to the issue I had open.

<ChrisL> http://wiki.csswg.org/spec/css2.1#issue-198

<Bert> http://lists.w3.org/Archives/Public/www-style/2010Sep/0674.html

<ChrisL> http://lists.w3.org/Archives/Public/www-style/2010Sep/0674.html

glazou: Can you summarize, Steve?

szilles: The problem is that, with runins, floats inside the runins - how do they relate to floats that precede and follow the runin?
... I think we decided in Oslo that runins that run in are rendered/treated as a child of the block. So a float that originally came between the runin and the block is now treated as preceding the runin.

glazou: MS, do you think run-in is sufficiently bad to be at risk?

arronei: runins are pretty much fully implemented in IE and Opera. A few iffy areas around tables, but otherwise the 100+ tests we have are passed by both, and by Safari for many of them.

glazou: So, Elika, is it worth putting it at risk?

fantasai: Eh, if we have interop in everything but floats, that's probably good enough for most people.

szilles: That's separate from the issue at hand, though, about runin and float interaction.

glazou: If it ends up being at risk, we can just mark it undefined.

Bert: I have two things to say.
... First is that the "first line" bit is complicated.
... It's not wrong, it's just unclear I think.

szilles: I'm okay with juts saying "rendered as an inline (refer to section XX)".

Bert: Second was a typo.

<glazou> we can't hear you

szilles: There were comments from bz, but they were just saying that the general area about how floats are painted is undefined, but the actual part I defined is fine.

<glazou> type here pls

<ChrisL> Its is better to me to have normative text included and to put the subfeature at risk

<ChrisL> rather than the whole feature

<ChrisL> or having nio definition and thus no guidance for implementors

<glazou> yes

<bradk> agreed

RESOLUTION: Accept szilles' edits regarding floats and runins, but mark the float/runin interaction as at-risk.

glazou: Next agenda item would be about CSS3 Color, but tantek isn't on the call and we can't hear Chris.
... We'll skip it and come back.

<ChrisL> noo

glazou: Next is about the style attribute.
... Bert, I see you sent a request for PR, anything else need to happen?

<glazou> ChrisL: isn't it YOUR phone ?

Bert: No, we just need to wait for the telcon.

glazou: next is about epub laiason. Bert?

Bert: w3c has a laiason with the epub group. It's handled by MikeSmith on this side.
... They've asked us to provide more laiason with CSS.

<glazou> plinss_: could you conference ChrisL in ?

Bert: They've shown their schedule for their expected spec, and they seem to be in quite a hurry to get it published.

<glazou> ok ChrisL

jdaggett: If you dig around in their doc, their set of requirements is fairly jaw-dropping - hyphenation, encoding schemes, etc.
... So, what is the goal?

Bert: Good question. I don't have one. I'm asking for how best to coordinate with them; we should also discuss what our goal for coordinating with them as well.
... We should probably tell them that their goal is too high - we can't finish that much of CSS3 in time, but we can do parts of it.

howcome: hyphenation is finished, for example.

Bert: We have good proposals, can they be brought to CR in time?

howcome: I think so, if we fork it out separately.

fantasai: I think I agree with howcome that we can get hyphenation to CR. I'm also trying to get the rest of CSS3 Text to CR, because many of the features they want are in that draft.

Bert: jdaggett, you said there were more features?

jdaggett: Vertical text - we're still arguing about the fundamental properties. It needs to be well-defined by the end of the year.
... To meet the epub schedule, at least.

fantasai: We will have two impls by the end of the year.

jdaggett: Impls of what?

fantasai: what Antenna House and what Hyatt wrote for Webkit will be compatible.

jdaggett: What have we agreed upon?

fantasai: AH and Hyatt have implemented logical margins.

howcome: That doens't mean that's what should be defined.

jdaggett: Every time we discuss this we get to a point, but don't put things into a spec.

szilles: What we agreed in Oslo was to see what your proposal was and then discuss it at TPAC.

fantasai: Yes.

Bert: I'd like to go back to coordination with IDPF. Koji, what's your role in there?

kojiishi: I'm willing to take over what Shinyu has been doing.
... If you have anything else you expect the IPDF laiason to do beyond what Shinyu has been doing, I can do that.

Bert: You plan to participate in both our meetings and IPDF meetings?

kojiishi: Yes.
... EGLS is a globalization effort within IPDF. I talk with him often, but I'm otherwise not very familiar with it.

<jdaggett> no ???? = EGLS

kojiishi: i can talk to my contact in IPDF for what they really expect.

howcome: I think something that might help is to have a set of use-cases.
... I think what EPUB is looking for is perhaps simpler than the problem we're seeing on our side.

kojiishi: Vertical text and globalization are on our radar. Hyphenation I'm not tracking right now.

howcome: That's fine. I think the biggest issue we have is vertical text. If we could get proposed use-cases - typical things they need to solve - then we could more easily find a solution for that.

<kojiishi> http://code.google.com/p/epub-revision/wiki/EGLS_solutions

szilles: There is a requirements doc that epub is putting out.

kojiishi: They haven't finished the prioritization quite yet. They should have it be their meeting on Oct 5th.
... I'll be attending that meeting.
... Hopefully we'll find out what we're missing and what items are high priority.

szilles: I believe Adobe will also have a rep on that meeting, though not me.

Bert: Would it be possible to have reps come to our meeting in Lyon in November?

glazou: That would be fine with me.

Bert: That we could probably ask via MikeSmith.

<scribe> ACTION: Bert to Ping MikeSmith about getting IPDF people at TPAC. [recorded in http://www.w3.org/2010/09/29-CSS-minutes.html#action01]

<trackbot> Created ACTION-268 - Ping MikeSmith about getting IDPF people at TPAC. [on Bert Bos - due 2010-10-06].

glazou: We need a laiason; the deadlines of CSS and EPUB don't quite match; we need a closer discussion with EPUB people to really decide something.

ChrisL: Are the EPUB deadlines hard, or are they just a guess?

szilles: I believe they're actually trying to operate toward those deadlines.
... EPUB primarily wants to base their work on existing standards.
... There was something that Sharp and related companies put together a while back that they're pushing to use.
... The real problem is having something in the near term that would better align with w3c's tech than some of th ealternatives.
... That's partly why teh deadlines are so short.
... There are certainly a substantial number of people in the EPUB discussion that would prefer an approach that was closer to w3c standards, if they could.

bradk: Can they align with working drafts, perhaps with an epub prefix? Or do they have to have PRs?

<ChrisL> if they dont get the featurs they need from w3c specs would they drop the feature, extebd the deadline, or pick a different source of standards?

szilles: If they like what they think they're going to get, they may accept them before CR, because there's really nothing else out there right now except ISO standards.

kojiishi: I think generally they need CR.

<kojiishi> http://code.google.com/p/epub-revision/wiki/CSS3Relations

kojiishi: As far as I know they haven't decided yet, and they'll be discussing this in Taiwan on Oct 5th.

szilles: So, I believe that Elika and friends are tyring to produce a do we can look at and discuss by TPAC, to see if we have neough consensus to move ahead.

jdaggett: There are two specs here - CSS3 Text and CSS3 Writing Modes.

szilles: Elika, you were doing Text, not Writing Modes, right?

fantasai: Right; I've made a lot of edits to Writing Modes, but it's not ready to publish right now. CSS3 Text is ready to publish as a WD.
... The current CSS3 Text draft is quite outdated. The new draft has addressed nearly all the issues raised with it.
... I'd like to get a new draft published before the Taiwan meeting next week.

<jdaggett> jdaggett: i'm fine with publishing a new draft of CSS3 Text

<glazou> 30 seconds before next item plase

<glazou> please

szilles: I haven't looked at the new Text. Has it changed?

<tantek> also for publishing new draft of CSS3 Text

fantasai: Yes, a lot. [lists changes]

<glazou> TIME

<jdaggett> glazou: let's resolve to publish and move one

<kojiishi> http://dev.w3.org/csswg/css3-text/

RESOLUTION: Publish a new WD of CSS3 Text.

<tantek> ChrisL - maybe this: http://dev.w3.org/csswg/css3-text/

glazou: Now, CSS3 Color.

<ChrisL> http://dev.w3.org/csswg/css3-color/issues-lc-2008.html

<tantek> Thanks much Chris!

ChrisL: I spent today making a Disposition of Comments. It's not in a great shape.
... We need what the WG did, what the commenter responded, etc.
... We have at least 3 where the commenter said they weren't happy, a bunch where they didn't respond, and 1 from the XSL WG where we didn't respond to them at all (we made edits, but didn't tell them).
... We mostly agreed with them, which makes it easy.
... I sent out a few others asking for confirmation.

<glazou> lol

tantek: The DoC is looking good, thanks for writing it up.

ChrisL: I didn't get the biblio updated, but I will.

tantek: Did you see the most recent followup regarding scRGB?

ChrisL: There was a recent discussion about adding a bunch of ICC spaces.
... Should we treat it as an LC comment or what?

tantek: I think it's out of scope right now. We should reject it for now, and have it as possible functionality for CSS4 Color.

<glazou> sylvaing: lol

tantek: Does the DoC look good otherwise?

ChrisL: The XSLWG response we should probably wait a week or so for.

tantek: So we're pretty happy with the DoC, and we'll give commenters a week or two to respond. What's next?

ChrisL: Then we produce a director's ???, showing that we exit CR correctly.

glazou: Any normative dependency on CSS2.1?

ChrisL: yes.

glazou: So it'll be at the same state as CSS3 Selectors?

ChrisL: Yeah. It precludes PR, but not Rec.

glazou: So yeah, we need to push CSS2.1 quick.

ChrisL: Selectors already has "provisional Rec" contingent on CSS2.1 going to Rec.

<glazou> View mode review request for email or next telcon

<Ms2ger> ...Adjourned?

<glazou> yes

Nope, simultaneous mass murder.

<glazou> lol

<fantasai> lol

<glazou> TabAtkins_: saw my private msg?

glazou: yeah.

<tantek> TabAtkins_ to be clear it does NOT preclude PR

<tantek> it does preclude REC

<tantek> (per minutes above)

Argh, I reversed them.

fantasai: Could you fix that when you prepare minutes?

<tantek> glazou - did we record an action for Bert to setup a call with the Director?

<ChrisL> s/It precludes PR, but not Rec./It precludes Rec, but not PR

<tantek> or are we doing that later?

<Bert> better wait until we actually have the DoC ready.

<Bert> (And maybe Chris and do it himself? That saves me work :-) )

<ChrisL> bert, yes I can do it once DoC is ready

<tantek> Bert, ChrisL - excellent - thanks much!

<tantek> ChrisL - how much time did you want to give commenters to follow-up? is a week sufficient or would you be more comfortable with 2 weeks?

<ChrisL> I think a week is reasonable. Willtalk with xsl fo staff contact

<tantek> ok

<fantasai> RRSAgent: make minutes

<fantasai> RRSAgent: make logs public

<fantasai> RRSAgent: make minutes

- DRAFT -

SV_MEETING_TITLE

29 Sep 2010

Attendees

Contents

Summary of Action Items

Scribe.perl diagnostic output