Re: [css-masking] Deprecating 'clip'

On 1/10/14, 9:20 AM, "Dirk Schulze" <dschulze@adobe.com> wrote:

>
>On Jan 8, 2014, at 6:25 PM, Tab Atkins Jr. <jackalmage@gmail.com> wrote:
>
>> On Wed, Jan 8, 2014 at 12:36 AM, fantasai
>><fantasai.lists@inkedblade.net> wrote:
>>> # Note: With this specification the clip property is deprecated.
>>> 
>>> Deprecation is a normative statement, so shouldn't be in a note.
>
>The sentence is not a note anymore.
>
>>> 
>>> # <shape>
>>> #
>>> # In CSS 2.1, the only valid <shape> value is: rect(<top>, <right>,
>>> <bottom>, <left>)
>>> 
>>> This isn't the CSS2.1 spec.
>
>I changed the sentence. It does not mention CSS2.1 anymore.
>
>>> 
>>> Also, <shape> seems overly broad for something that expands only
>>> to rect(). I think we should change this type's name here and in
>>> CSS2.1 to something else (<clip-rect>?) and allow Basic Shapes
>>> to define <shape> for use everywhere else. It's very clumsy for
>>> <shape> to only define rectangles defined by two points and
>>> <basic-shape> to have much broader expressiveness than <shape>.
>> 
>> Agreed.  Note that with Bikeshed, you can refer to a function in a
>> grammar by using the <<foo()>> shortcut syntax, so there's no need for
>> us to define a grammar production at all for just rect().
>
>I removed the definition of <<shape>> and added a definition for
><<rect()>>. It is now up to CSS Shapes to pick up the term <<shape>> and
>I’ll use it in CSS Masking as well.

I would rather keep <basic-shape> in CSS Shapes, as they are less
expressive than other shape sources such as images.

Removing <shape> from the clip definition in masking is enough to solve
the minor issue Fantasai noted, I think.

Thanks,

Alan

Received on Friday, 10 January 2014 18:14:54 UTC