W3C home > Mailing lists > Public > www-style@w3.org > August 2009

Re: [CSSWG] Minutes and Resolutions 2009-08-12

From: Tab Atkins Jr. <jackalmage@gmail.com>
Date: Fri, 14 Aug 2009 15:44:36 -0500
Message-ID: <dd0fbad0908141344ic837576xf4f2e2cc2d0e9cec@mail.gmail.com>
To: Brad Kemper <brad.kemper@gmail.com>
Cc: fantasai <fantasai.lists@inkedblade.net>, "www-style@w3.org" <www-style@w3.org>
On Fri, Aug 14, 2009 at 3:11 PM, Brad Kemper<brad.kemper@gmail.com> wrote:
> On Aug 14, 2009, at 12:47 PM, fantasai <fantasai.lists@inkedblade.net>
> wrote:
>> Tab Atkins Jr. wrote:
>>>
>>> On Fri, Aug 14, 2009 at 2:02 PM, fantasai<fantasai.lists@inkedblade.net>
>>> wrote:
>>>>
>>>> Robert O'Callahan wrote:
>>>>>
>>>>> One other thing...
>>>>>
>>>>> On Thu, Aug 13, 2009 at 10:40 AM, fantasai
>>>>> <fantasai.lists@inkedblade.net
>>>>> <mailto:fantasai.lists@inkedblade.net>> wrote:
>>>>>
>>>>>   * sgalineau can definitely see gradients used for border-image
>>>>>
>>>>>
>>>>> I can't, really, at least not as-is. Nor do I see them as being useful
>>>>> for
>>>>> 'list-style-image', 'cursor' or 'content'.
>>>>>
>>>>> Orthogonality is cool, but actual use-cases beyond 'background' seem
>>>>> pretty thin. If we have trouble specifying the behaviour of gradients
>>>>> for
>>>>> those other properties (because there's no natural size to use for the
>>>>> gradient, perhaps), I think we could just disallow them.
>>>>
>>>> Sounds like a case for making <gradient> separate from <image>.
>>>
>>> As Brad mentioned, though, we already have *images* without intrinsic
>>> dimensions in SVG.  How do we deal with those?  Are they unsuitable
>>> for use in list-style-image, etc?
>>> Whatever solution is used for SVG without intrinsic dimensions, an
>>> identical solution should be applicable to gradients.
>>
>> Behavior for that, and cursors, and background images, and
>> list-style-image
>> is all in the CSS2.1 spec. (IIRC Anne filed the issue a year or two ago.
>> ;)
>> So we could just put it in <image>. But I can see that in several places
>> we'd want slightly different behavior for gradient(), or want to allow
>> it in places where standard images are not allowed (e.g. border-color).
>
> How about a new property, border-overlay, that just stretches a border to
> the border box, and shows the part that intersects with the border. Done,
> end of definition. ;)
>
> Or 'border-background' that does everything that background does (except
> background-clip), but just within the border?

You can do that currently by specifying multiple backgrounds, with the
first set to border-box and the second set to padding-box.

> That seems more like a glow or a box-shadow.

Hmm, you're right; border-colors, at least in many simple gradient
cases, can be replaced by box-shadow just fine.

> It's not what I imagined for a gradient border. I imagined something that did not ignore the direction. You can have a nice bezel effect if, for instance, your border gradient does dark to light while your padding box goes light to dark at the same angle.

::points up to his previous comment about how to achieve that effect::

~TJ
Received on Friday, 14 August 2009 20:45:31 GMT

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