Re: [whatwg] style scoped available as experimental feature, discussing @-rules

[Ian Hickson]

On Wed, 7 Mar 2012, Roland Steiner wrote:
 
 In this regard, I would also like to try to re-ignote the discussion on 
 the fate of @-rules within style scoped: As has been discussed on 
 these mailing lists previously, as well as informally, implementing a 
 scoped version of @font-face is quite difficult, and there are several 
 voices against even attempting it. Consequently, without an 
 implementation of @font-face, it may make more sense to also ignore 
 @keyframes and @region for consistency reasons rather than have a 
 mix-and-match, even though the latter 2 shouldn't be that hard to 
 implement.
 
 What are your thoughts on this?

I think not supporting @font-face would suck, but not enough to destroy 
the feature. Not supporting the others seems more serious. If nobody 
implements them, then the spec will eventually say they aren't to be 
implemented. If anyone ever does implement them, then they'll be something 
to implement for compatibility.

Generally though it seems lame to not support them. :-)

-- 
Ian Hickson   U+1047E)\._.,--,'``.fL
http://ln.hixie.ch/   U+263A/,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'

[whatwg] style scoped available as experimental feature, discussing @-rules

[Roland Steiner]

Hi all,

First, sorry for the cross-post, I didn't know which mailing list was most
appropriate, and style scoped has previously been discussed on both.

Now, a shiny-new implementation of style scoped has made it into WebKit
and is now available as an experimental feature behind a run-time flag in
Chromium. See also the html5rocks article that has more details about it:
http://updates.html5rocks.com/2012/03/A-New-Experimental-Feature-style-scoped
. As
the main developer of the WebKit implementation, I would hereby like to
solicit feedback on the feature and real-world experiences. :)

Currently, there are no known bugs, the :scope pseudo class is implemented,
but the following @-rules not yet: @font-face, @keyframes, @region, @global.

In this regard, I would also like to try to re-ignote the discussion on the
fate of @-rules within style scoped: As has been discussed on these
mailing lists previously, as well as informally, implementing a scoped
version of @font-face is quite difficult, and there are several voices
against even attempting it. Consequently, without an implementation of
@font-face, it may make more sense to also ignore @keyframes and @region
for consistency reasons rather than have a mix-and-match, even though the
latter 2 shouldn't be that hard to implement.

What are your thoughts on this?


Cheers,

- Roland