W3C home > Mailing lists > Public > www-style@w3.org > November 2013

[CSSWG] Minutes Telecon 2013-11-27

From: Dael Jackson <daelcss@gmail.com>
Date: Thu, 28 Nov 2013 06:56:30 -0500
Message-ID: <CADhPm3tP8wt1OOTjcXoiNhZXv7gKnZbVfNx2viEBrGbpSU9vWg@mail.gmail.com>
To: www-style@w3.org
Shapes
------

  - RESOLVED: Last Call for Shapes.
  - Deadline is 7 January, 2014.

WebVTT
------

  - Everyone was requested to review the WebVTT spec for ::cue and
         ::cue-region for discussion in two weeks.

Dates for Seattle F2F
---------------------

  - RESOLVED: Seattle F2F dates are 27-29 January, 2014

Remove nav-down/-up/-left/-right from CSS 3 UI at-risk list
-----------------------------------------------------------

  - RESOLVED: Remove nav-up/down/left/right from CSS3 UI at-risk list.

Spread Radius Formula
---------------------

  - Two potential ways to make border animation smoother were either the
         cubic function created by fantasai and leaverou [1 + (x-1)^3]
         or a bezier function.
  - RESOLVED: Use cubic function for spread radius bending.
  - Publishing a Last Call for Backgrounds was also discussed with the
         group deciding to address some issues on the mailing list and
         publish after next week's call.

CSSOM
-----

  - RESOLVED: Publish new WD of CSSOM with improved changes list.

Grids
-----

  - Rossen will look over the issues raised by SimonSapin and respond on
         the mailing list.

Charter
-------

  - Everyone was requested to look over the updates from Bert for next
         week's call.

=====FULL MINUTES BELOW======

Present:
  David Baron
  Bert Bos
  Bruno de Oliveira Abinader
  Elika Etemad
  Sylvain Galineau
  Daniel Glazman
  Israel Hilerio
  Koji Ishii
  Dael Jackson
  Philippe Le Hégaret
  Peter Linss
  Christopher Palmer
  Anton Prowse
  Simon Sapin
  Alan Stearns
  Lea Verou (IRC only)

Regrets:
  Tab Atkins
  Dave Cramer
  Justin Erenkrantz
  Brad Kemper
  Kang-Hao Lu
  Florian Rivoal
  Dirk Schultz
  Leif Arne Storset
  Stephen Zilles

Shapes
------
  ScribeNick: fantasai

  <astearns> http://dev.w3.org/csswg/css-shapes/
  astearns: I made edits so shapes is ready for LC.
  fantasai: Sounds good to me.
  glazou: Me too.
  Rossen: Let's do it.

  RESOLVED: LC for Shapes.

  Bert: Which groups shall we ask for review?
  astearns: SVG certainly.
  glazou: DigiPub IG.

  [discuss LC period]
  [deadline: Tuesday, 7 January 2014 with the understanding that people
             can ask for extension.]

WebVTT
------

  glazou: plinss mentioned request from WebVTT
  plinss: I forgot to forward their e-mail to the WG...
  plinss: They want us to look at ::cue-region etc.
  <plinss> https://dvcs.w3.org/hg/text-tracks/raw-file/default/608toVTT/region.html

  glazou: How much time to we have?
  plinss: They didn't mention it in the e-mail.
  <plinss> http://dev.w3.org/html5/webvtt/#css-extensions

  Bert: I'm interested in reviewing
  glazou: Let's shoot for 2 weeks
  ACTION everyone: Review the WebVTT spec for ::cue and ::cue-region
  * RRSAgent records action 1
  <trackbot> Error finding 'everyone'. You can review and register
             nicknames at <http://www.w3.org/Style/CSS/Tracker/users>.

  glazou: I've another concern about CSS stuff being in other specs
  glazou: Shadow DOM is adding CSS stuff to Shadow DOM spec
  <glazou> https://www.w3.org/Bugs/Public/show_bug.cgi?id=23467
  glazou: There's an experimental feature introducing two new
          combinators to Selectors.
  glazou: We should monitor this, before it becomes too late to discuss
          anymore.

  <abinader> I'd like to mention if is it possible to add CSS3 Text
             Decoration properties as CSS extensions of WebVTT.
  <abinader> Currently only text-decoration is listed there,
  <abinader> But on CSS3 spec it is now a shorthand.
  <fantasai> File that against webVTT
  <abinader> fantasai: right

  <fantasai> That's TabAtkins's summary of Shadow DOM stuff:
             http://dev.w3.org/html5/webvtt/#css-extensions

  <SimonSapin> About Grid, I'd like to hear from Rossen on
         http://lists.w3.org/Archives/Public/www-style/2013Nov/0382.html

Dates for Seattle F2F
---------------------

  glazou: Dates are firm?
  sylvaing: We currently have that week (27 Jan - 31 Jan) booked.
  sylvaing: W can swap CSS/SVG, but everything is booked for that week
  sylvaing: Right now CSS is Monday - Wednesday with a joint SVG meeting
            on Wednesday.
  glazou: Maybe we should freeze that so people can book their flights.

  <dbaron> I probably prefer CSS first (i.e., M-W as currently planned),
          because I'd like to be in the bay area at least a few days
          that week, and M-W lets me return Wednesday night.

  RESOLVED: Seattle F2F dates are 27-29 January, 2014

  sylvaing: I will add more info next week to the planning wiki.
  glazou: I got a request from szilles to sort out September dates
  fantasai: Maybe we should sort this out on a call with more people.
  * sgalineau didn't hear whether we had a location for september...
  <plh> Sylvain, location is Sophia.
  <sgalineau> Cool, thanks.
   * dbaron is not crazy about meeting in Sophia...

  ACTION glazou: send email wrt dates for September F2F
  * RRSAgent records action 2
  <trackbot> Created ACTION-600 - Send email wrt dates for september f2f
             [on Daniel Glazman - due 2013-12-04].

Remove nav-down/-up/-left/-right from CSS 3 UI at-risk list
-----------------------------------------------------------

  glazou: We have 2 or 3 interoperable implementations at this time,
          implementing the four properties.
  glazou: I contributed 64 tests that need review.
  glazou: The tests are partly based on what Opera did.
  glazou: The good news is that all tests are passed by at least 2
          implementations.
  glazou: So I don't think the properties are at-risk anymore, but
          that's up to WG to decide.

  sylvaing: How comprehensive is 64?  Is that 64 out of 200, 100...?
  glazou: I made a test for every normative statement in the spec.
  glazou: I think 64 is final.
  sylvaing: Or very close. Good.
  fantasai: I don't see a problem.
  sylvaing: Me neither.

  plh: What are the two implementations?
  glazou: Opera TV emulator is one.
  glazou: 2nd/3rd are Smart TV emulator from Samsung and build of
          Chromium that matches that implementation.
  glazou: The emulator is difficult to test,
  glazou: But we can use the standalone browser Chromium.
  glazou: It's the same feature based on same code.
  rossen: First one is Presto?
  glazou: Yes, Presto.
  Rossen: So we have one Presto and one Blink.

  RESOLVED: Remove nav-up/down/left/right from CSS3 UI at-risk.

  ACTION glazou: Make that edit
  * RRSAgent records action 3
  <trackbot> Created ACTION-601 - Make that edit [on Daniel Glazman -
             due 2013-12-04].

Spread Radius Formula
---------------------

  <glazou> http://lists.w3.org/Archives/Public/www-style/2013Nov/0290.html
  fantasai: At TPAC we were talking about the discontinuity of the
            spread shape between border radius of zero and non-zero.
  fantasai: And dbaron suggested making some kind of gradual increase.
  fantasai: So after some informal discussions, we came up with a fairly
            simple formula,
  fantasai: 1 + (x-1)^3
  fantasai: Which describes how much of the spread distance to add to
            the curve radius while it is less than the border radius.
  fantasai: This gives the same results in case of zero and in case of
            border-radius greater than spread.
  fantasai: But in between it interpolates smoothly,
  fantasai: Slowing down the rate of change so that it hits constant 1
            smoothly,
  fantasai: And the rate of change at the beginning is not too fast.
  fantasai: The result looks better than the previous definition in all
            cases.
  fantasai: Because it used to look weird when you had a small
            border-radius and large spread radius.

  <astearns> x = border-radius/spread
  <astearns> also used for calculating the margin-box shape, where x =
             border-radius/margin

  <glazou> see http://is.gd/miUFlA
  <fantasai> http://dev.w3.org/csswg/css-backgrounds/spread-radius
  fantasai: You can play with formulas we tried there.

  plinss: I added a bezier function to that demo as well.
  plinss: It adds a lot of ability to change curve.
  plinss: Cubic-bezier is what's used in transitions.
  fantasai: I think the cubic looks good enough, but if people have
            reasons to tweak the behavior (moves too fast here, too slow
            there), you can play with cubic bezier
  <Rossen_> ^3 ftw
  <astearns> +1 for cubic from me
  * sgalineau likes cubic best from the demo page
  * sgalineau though the ability to use a bezier could be fun in a
              future level...
  <glazou> sgalineau Agreed.
  <glazou> sgalineau We used to discuss names, now we discuss maths :-)

  fantasai: My suggestion is to accept the cubic function in the spec,
            and take it to LC.
  glazou: Looks good to me.
  Bert: I see little difference between them.
  fantasai: You have to set the spread to something large, then bend the
            radius very slowly from zero to equal to spread.
  fantasai: The difference is in the rate near zero.
  fantasai: Cubic looks better.

  glazou: Do we want to extend functionality in the future?
  dbaron: I don't think is something we want multiple options for.
  sylvaing: Might want the possibilities, but for now cubic is great.
  <SimonSapin> +1 for cubic, but this looks to me like a tiny difference
               that you have too for. Let's pick something and move on.

  RESOLVED: Use cubic function for spread radius bending.

  fantasai: I want to take css3-background through LC.
  astearns: The spec doesn't currently describe how background-position
            animates correctly.
  astearns: I can post that to the list.
  fantasai: [Reads off Animatable line of spec]
  fantasai: That's clearly wrong.

  dbaron: I think a <position> can be represented as a pair of values if
          you use calc() functions.
  dbaron: You can represent all positions as two values of length,
          percentage, or calc.
  dbaron: That's what we do in Gecko.
  dbaron: But you also can't animate between that purely in the computed
          value space.

  astearns: In Gecko you might do that, but the Animatable line doesn't
            match the computed value.
  fantasai: So, the reason the computed value line is written that way
  fantasai: Is because in L4 we expect to add logical keywords.
  fantasai: And you cannot animate between physical and logical
            coordinates.
  fantasai: So, we could make physical coords animatable like in Gecko,
  fantasai: And for logical coords, you'd do the same but with keywords
            for start start origin,
  fantasai: And then you can't animate between the two.
  dbaron: I don't think we should do this during the call.

  dbaron: So I think this spec needs to write a definition for animating
          <position>.
  fantasai: Would that go in this spec, or values and units?
  dbaron: Whatever defines <position>.
  glazou: So, should we wait for that before LC?
  dbaron: Could probably handle as LC comment.

  [discussion about publishing timelines]

  fantasai: We'll plan to publish Thursday, send in publication request
            early, and just prep the draft after the call next week.
  glazou: Anything else on that topic?
  fantasai: astearns, you'll send in the e-mail to publish?
  astearns: Yes.

CSSOM
-----

  <glazou> http://dev.w3.org/csswg/cssom/
  fantasai: I think it's pretty out of date. We should publish an update
  <SimonSapin> +1 to publish
  <plh> +1 to publish
  Bert: I agree in principle, but what is state of current editor's
        draft?
  Bert: Is there a list of changes since previous draft?
  Bert: I don't object to publishing, but was wondering if anyone could
        tell me what actually changed since last time.
  glazou: I have the same comment, don't object to publishing, but would
          also like to know what changed.

  RESOLVED: Publish new WD of CSSOM with improved changes list.

Grids
-----

  SimonSapin: As discussed with Tab on the mailing list,
  SimonSapin: Syntax of grid-template-areas.
  SimonSapin: We would like hear from microsoft in particular.

  SimonSapin: String values inside that property are "parsed as CSS",
  SimonSapin: We end up with possibilities of double-escaping.
  SimonSapin: Tab and I agree it would be better to define a small
              special parsing
  SimonSapin: For content of templates inside string,
  SimonSapin: i.e. how to parse into identifiers.
  Rossen: OK, I will go over today and send email. Don't think we need
          to discuss over call

Charter
-------

  glazou: Another item I'd like to discuss, maybe not today, is charter.
          See message from Bert.
  glazou: Please review charter and send comments for next week.

[Meeting closed]
Received on Thursday, 28 November 2013 11:56:57 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 28 November 2013 11:57:00 UTC