W3C home > Mailing lists > Public > www-style@w3.org > February 2012

Re: [css3-transforms] Extending CSSOM interface

From: Dirk Schulze <dschulze@adobe.com>
Date: Fri, 17 Feb 2012 09:57:59 -0800
To: Aryeh Gregor <ayg@aryeh.name>
CC: "L. David Baron" <dbaron@dbaron.org>, www-style list <www-style@w3.org>
Message-ID: <6929DB26-511C-47B5-8C0B-63D957DA0BC0@adobe.com>
I still believe that the interfaces for CSS Transforms are really useful. And I would be happy to work on that and see it for CSS Transforms. But it feels like these changes are not widely accept because of following reasons:

* People are afraid that it could block CSS3 Transforms to reach CR soon.

* Some discrepancies how to deal with DOM2 (getPropertyCSSValue() and CSS*Value) and how to improve CSSOM API.

Maybe we can find a compromise:

Remove CSSTransformValue and CSSMatrix from the current spec like proposed by Simon Fraser, Boris Zbarsky and Aryeh (who was unsure about CSSMatrix).

Work on an interface (CSSTransformList, CSSTransform(Value), CSSMatrix or what ever) for CSS4 Transforms. This interface would give authors direct access to modify transformation functions.

Greetings,
Dirk

On Feb 17, 2012, at 8:20 AM, Aryeh Gregor wrote:

> On Thu, Feb 16, 2012 at 5:37 PM, Dirk Schulze <dschulze@adobe.com> wrote:
>> Dropping CSSTransformValue and CSSMatrix? Or just the first one?
> 
> CSSMatrix has a different problem.  We probably want a single class
> for matrices that's shared by SVG, CSS, WebGL, etc.  I don't think we
> want separate classes for SVGMatrix, CSSMatrix, etc.
> 
> On Thu, Feb 16, 2012 at 7:05 PM, Dirk Schulze <dschulze@adobe.com> wrote:
>> Even if CSSOM is in development, there is a W3C recommendation that specifies getPropertyCSSValue():
>> 
>> http://www.w3.org/TR/DOM-Level-2-Style/css.html#CSS-CSSStyleDeclaration
> 
> However, nothing in the transforms spec says that
> getPropertyCSSValue('transform') should return a CSSTransformValue.
> If it said that, then the interesting parts would actually be
> testable.  That would be a step forward -- although if nobody but
> WebKit wants to implement it, which so far seems to be the case, it
> should still be removed from the spec IMO.
> 
> On Thu, Feb 16, 2012 at 7:27 PM, L. David Baron <dbaron@dbaron.org> wrote:
>> There should be an open issue on removing this material from the
>> transforms spec.  Either that, or it's possible (since I have a
>> vague memory of a discussion) that we already resolved to remove
>> this material from the spec.  We shouldn't be specifying anything
>> new that depends on the CSSValue interface.
> 
> Well, I filed a bug: https://www.w3.org/Bugs/Public/show_bug.cgi?id=16003
Received on Friday, 17 February 2012 17:58:33 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:50 GMT