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

Re: [css3-images] element() "ignoring" transforms

From: Dirk Schulze <dschulze@adobe.com>
Date: Sat, 10 Mar 2012 06:57:27 -0800
To: "Tab Atkins Jr." <jackalmage@gmail.com>
CC: "L. David Baron" <dbaron@dbaron.org>, Řyvind Stenhaug <oyvinds@opera.com>, "www-style@w3.org" <www-style@w3.org>
Message-ID: <55EF5E10-8758-4E66-A6C0-7817515746FF@adobe.com>

On Mar 5, 2012, at 11:28 AM, Tab Atkins Jr. wrote:

> On Mon, Mar 5, 2012 at 10:53 AM, L. David Baron <dbaron@dbaron.org> wrote:
>> On Monday 2012-03-05 10:46 -0800, Tab Atkins Jr. wrote:
>>> On Mon, Mar 5, 2012 at 9:47 AM, Řyvind Stenhaug <oyvinds@opera.com> wrote:
>>>> Regarding the eleemnt() function, CSS3 Images says
>>>> "If the referenced element has a transform applied to it or an ancestor,
>>>> the transform must be ignored when rendering the element as an image.
>>>> I think it's probably worth detailing what it means for a transform to be
>>>> "ignored". Should it be as if the transform property were 'none'? Note
>>>> that this would have some additional effects, since a transformed element
>>>> normally acts as a containing block for fixed-position descendants, for
>>>> instance (even if it's just an identity transform).
>>> I should specify that.  The intent is similar to how SVG defines this
>>> kind of thing; that a "virtual" transform that's the inverse of the
>>> CTM is applied after the element's transform.  Thus, the side-effects
>>> you get from transforming the element (like becoming a fixpos
>>> container) are still preserved.
>> I'm not sure this is the right way to specify this:  inverting has
>> different results if the element has or is inside an element with a
>> sigular transform.  And I think element() should be able to capture
>> something that's inside a singular transform (despite that that
>> transform makes the original not show up).
> Hmm, indeed.  I can weasel my way around that, since the "virtual
> transform" isn't actually visible anywhere.
Just as a reference: CSS Transforms allows transforms for child elements like patterns or masks. Would be great if both specs can align the description.


> ~TJ
Received on Saturday, 10 March 2012 14:58:12 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 11 February 2015 12:35:06 UTC