- From: fantasai <fantasai.lists@inkedblade.net>
- Date: Wed, 23 Jun 2010 12:00:57 -0700
- To: "www-style@w3.org" <www-style@w3.org>
Summary:
- Reviewed CSS2.1 test suite status
- Reviewed open CSS2.1 issues
- Discussed whether preformatted text should be allowed to justify.
(CSS2.1 Issue 53)
- RESOLVED: Proposal accepted for CSS2.1 Issue 107 (top/left/bottom/right computed values)
- RESOLVED: Proposal accepted for CSS2.1 Issue 145 (Handling forced line breaks and bidi)
- RESOLVED: Proposal accepted for CSS2.1 Issue 149 (Definitions of px and physical units)
- RESOLVED: Proposal accepted for CSS2.1 Issue 151 (Allow XML to use preshint level)
- RESOLVED: Proposal accepted for CSS2.1 Issue 160 (Explain whether :first is :left or :right)
- Accepted in principle to change spec for CSS2.1 Issue 172, but need exact wording.
====== Full minutes below ======
Present:
Tab Atkins
David Baron (late)
Beth Dakin
Elika Etemad
Simon Fraser
Sylvain Galineau
Daniel Glazman
Brad Kemper
Peter Linss
Steve Zilles
<RRSAgent> logging to http://www.w3.org/2010/06/23-CSS-irc
Scribe: Tab Atkins
CSS2.1 Test Suite
-----------------
glazou: Elika, test suite status?
fantasai: Status is we published alpha 3 last week. There's a bunch
of moz tests that still need to be imported. Arron says
he's fixed the ones that were wrong, but there was still
some errors when I checked yesterda.
fantasai: I'll try and publish beta 1 next week, and we'll ask people
to review them and make sure everyone is happy getting
implementation reports based on them.
fantasai: We plan to get implementation reports in before the F2F.
fantasai: We plan to publish 2 betas, on a 3 week cycle.
glazou: You tweeted something about reftests. Is that related?
fantasai: No, that was me working with moz stuff.
CSS2.1 Issues
-------------
glazou: We have roughly 20 issues that are still undiscussed or lack a proposal.
glazou: I think the first one up is issue 53.
http://lists.w3.org/Archives/Public/www-style/2008Jun/0227.html
<plinss> http://wiki.csswg.org/spec/css2.1#issue-53
glazou: fantasai, this is something you posted.
tabatkins: fantasai's proposal sounds great, that glyphs and whitespace
in pre aren't altered for justification purposes.
szilles: So that says that pre is more powerful than justification?
glazou: Yes.
fantasai: I don't quite understand why you can't justify pre-wrap,
since it wraps, but I'd be fine with something that just
said you can't justify pre.
szilles: I thought that if the spacing are there even with the wrap,
you might have columns that you don't want to go jagged.
<sylvaing> you don't want to mess up Bert's ASCII art. Or Hixie's cat.
szilles: It seems odd anyway to have both pre and justify on an element.
tabatkins: You may have an inline with pre embedded in a justified element.
http://lists.w3.org/Archives/Public/www-style/2010Jun/0457.html
^^^ proposed solution
fantasai: We usually write plaintext in monospace to get things to line
up. But if you're using pre-wrap, maybe you're just writing
paragraphs, and would be okay with justifying.
fantasai: pre-line is definitely wrong in the spec and should be fixed.
But pre-wrap, I'm not sure of.
fantasai: One thing we could say is that you're not required to justify
in preformatted elements.
fantasai: And then maybe in css3 text we can clarify it more precisely.
glazou: I like that.
bradk: Are the other text-align values okay?
tabatkins: According to the proposal on the table, no, you can't
right-align pre text.
tabatkins: Nm, I'm misreading.
szilles: But if I break lines in my email, you don't want to justify that.
fantasai: Justification only affects lines that are soft-wrapped.
glazou: Any particular objections to the proposal on the table?
tabatkins: fantasai has a possible objection for the pre-wrap text.
glazou: I'll ping Thunderbird people for an answer on what's acceptable
to them, since it may affect email.
ACTION Daniel: Ping Thunderbird for their input on how the changes to
the spec around pre/pre-wrap and justification may
affect email.
fantasai: We can just say that justification isn't required.
tabatkins: I'd prefer not having it be impl-defined.
sylvaing: If the impls are disagreeing today, isn't it possible that
email clients right now are having problems?
tabatkins: Only if they're mixing it with justification, which I doubt
anyone is right now.
glazou: Next issue is 101, assigned to dbaron.
http://wiki.csswg.org/spec/css2.1#issue-101
Next one for the wg is 107.
http://wiki.csswg.org/spec/css2.1#issue-107
fantasai: The cases that this affects are where you're inheriting
top/bottom/left/right, which almost nobody does.
fantasai: The proposal is to correct the spec so that it's
consistent with impls.
glazou: And all impls do the same thing?
fantasai: Seems so; Arron and I ran a few testcases.
fantasai: And this is such a minor issue that I doubt it matters what
we actually do, so long as it's specified.
RESOLVED: Accept fantasai's proposal for resolving issue 107.
glazou: Next is issue 110, on Tab.
tabatkins: Haven't been able to write up a new proposal, sorry.
glazou: Next is issue 120.
fantasai: That'll take a while, since I have to go through the whole spec.
fantasai: There are like 4 issues on me that I need to deal with still.
glazou: Issue 138, on tab.
tabatkins: I wrote a proposal some time ago, I just didn't put it on
the wiki. I've done that now.
glazou: Everyone, read that and review for decision soon.
glazou: Next is issue 145.
tabatkins: I recall us discussing that during the BIDI meeting.
fantasai: Conclusion was not to change the CSS2.1 spec's handling
of forced line breaks.
fantasai: Proposal on the table addressing a clarification issue
around LS.
RESOLVED: Accept fantasai's proposal for resolving issue 145.
glazou: Next on the agenda is issue 149.
szilles: I don't see any possible way to have this resolve any other way.
glazou: At some point we have to decide, and I think the WG decided.
sylvaing: I think the only problem here is that Bert is the gatekeeper
of the spec.
...
fantasai: The issue we haven't discussed here is how this affects
Media Queries
fantasai: Frex, on an iphone with a width 2in media query, when does
it say yes/no?
sylvaing: For example, at first a webpage on the iphone is zoomed out
and full-size, but when you pinch and zoom in, you don't
want the media query to suddenly change, since the webpage
is still theoretically the same size.
glazou: Can we just resolve this on media queries?
sylvaing: If we accept the proposal now, we should go ahead and file
an issue on media queries, since it's in CR right now.
glazou: So, objections on the proposal?
* fantasai abstains from this discussion
RESOLVED: Accept fantasai's proposal to resolve issue 149.
NOTED: Issue on MQ about possible problems with Media Queries,
physical dimenions, and screen sizes.
glazou: Next issue, 151.
fantasai: SVG wants their presentational attributes treated as
presentational hints, not as ua stylesheet rules.
fantasai: But right now CSS defines otherwise, so they have to
explicitly overrule the CSS2.1 spec.
fantasai: This proposal alters the spec to allow XML languages to
have their presentational attributes either be on the
preshint level or the ua stylesheet level.
RESOLVED: Accept fantasai's proposal to resolve issue 151.
glazou: Next is issue 153.
fantasai: If you're aligning the baseline of a box to anything,
it's fairly well defined. But for any other alignment,
it's unclear which box of the element should be used.
fantasai: For replaced elements we have interop on margin boxes.
fantasai: But not so for inline non-replaced elements.
fantasai: I could try and spend more time on getting a proper
definition, but I don't know if we could impls in time
for CR.
* fantasai we should add a note that it would be defined in CSS3
<Zakim> +David_Baron
dbaron: I was just looking at IRC, and I don't think it needs to
be undefined.
dbaron: I think it might have been clearer in CSS1, but I think
it was well-defined somewhere.
ACTION dbaron: Produce a proposal for issue 153.
<glazou> current issue is 158
<tabatkins> http://lists.w3.org/Archives/Public/www-style/2010Jun/0124.html
<tabatkins> I think ^^^ is okay, but I'm still unsure of the exact
desired rendering of a particular edge case, so I'm not
sure if fantasai's suggestion fits or not.
<tabatkins> http://dbaron.org/css/test/2007/0329-blog-examples/1
<fantasai> I can dig up the emails from when we originally came up
with that issue
<fantasai> http://lists.w3.org/Archives/Member/w3c-css-wg/2007JanMar/0538.html
glazou: I suggest you two take the discussion offline and clarify this
between yourselves.
glazou: Next issue is 159, for elika.
fantasai: I haven't gotten to that yet.
<fantasai> http://wiki.csswg.org/spec/css2.1#issue-160
glazou: Issue 160. This should be easier to resolve.
glazou: The proposal seems to make perfect sense.
RESOLVED: Accept fantasai's proposal for resolving issue 160.
glazou: Issue 166, on fantasai.
<fantasai> Not done yet
glazou: Clarify different uses of "property value" and "component value",
and make a diff.
fantasai: Haven't been able to get to that yet.
glazou: Issue 167, backslash escapes.
glazou: So, where's the end of a stylesheet, if you "end" one with
a backslash?
tabatkins: There isn't interop, so whatever we decide should be okay.
tabatkins: I think it's just a matter of looking over Zack's proposed
changes and ensuring they really do have only the minimal
impact that he believes.
ACTION Daniel: Look over Zack's proposal for issue 167, ensure that it
doesn't have any unwanted effects.
glazou: Next issue, 170.
glazou: dbaron has listed 5 possibilities to resolve it.
dbaron: 6, counting the "leave it explicitly undefined" case.
bradk: Is there any interop on this?
ACTION Tab: Write some tests for issue 170, to see which option seems
closest to impls.
glazou: Next is issue 172.
<tabatkins> I don't have a problem with it. Table captions are rare in
the first place, and table captions that would overflow are
even more so.
<bradk> I'm fine with the proposal
ACTION fantasai: Propose wording for issue 172.
Meeting closed.
Received on Wednesday, 23 June 2010 19:01:38 UTC