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

Re: Feature queries

From: Andrey Mikhalev <amikhal@abisoft.spb.ru>
Date: Fri, 11 Dec 2009 19:43:01 +0300 (MSK)
To: Boris Zbarsky <bzbarsky@MIT.EDU>
cc: www-style@w3.org
Message-ID: <Pine.LNX.4.63.0912111914490.29183@master.abisoft.spb.ru>
On Fri, 11 Dec 2009, Boris Zbarsky wrote:

> On 12/10/09 4:44 AM, Andrey Mikhalev wrote:
>> don't understand.
>> syntax is the same. change of at-keyword cannot make syntax better.
>
> If @media is used, then extra syntax is needed to disambiguate the new thing 
> from media queries.  You end up with an extra level of parenthesis nesting, 
> etc.

'extra syntax needed' is not true, e.g.:
@media (height:600px) and (<cssproperty>[<:cssvalue>])
@media (min-height:600px) and (css-<cssproperty>[:<cssvalue>])

'extra level of parenhesis needed' is not true, e.g.:
@media (height:600px) and <func-css-supports>(<cssproperty>[:<cssvalue>])
@media (height:600px) and <func-supports>(text/css[,<cssproperty>[:<cssvalue>]])
@media (height:600px) and (text/css[,<cssproperty>[<:cssvalue>]])

>
>>> You're right that it would allow @import to be conditioned on this
>>> stuff; is that a highly desirable property?
>> 
>> you think it worth to load e.g. 1OOK of animated 3d transforms for 1K
>> document in phone UA over 10K link - before rejecting whole thing? ;)
>
> Maybe, if no one will in fact send 100K of 3d transforms with a 1K document.

at the moment (css2, without animations, transforms, advanced layout etc)
cnn.com ~ 20K content / 120K style:
22598  text/html http://www.cnn.com/
33916  text/css  http://i.cdn.turner.com/cnn/.element/css/3.0/common.css
24587  text/css  http://i.cdn.turner.com/cnn/.element/css/3.0/main.css
17465  text/css  http://i.cdn.turner.com/cnn/.element/css/3.0/connect/overlay.css
42548  text/css  http://i.cdn.turner.com/cnn/.element/css/3.0/personalization.css

> I understand what the use cases are for conditioning @import on this; I'm 
> asking whether there are serious plans to make use of such feature testing.

and i'm asking is it still css language or not. :)
if it is still css, it should provide _a way_ to achieve e.g. css design 
principle of network performance.

if not, will be great w3c explicitly claim and register this fact;
then your problem can be easily solved by upper protocols, e.g.
<link rel="stylesheet" type="text/css" href="pretty-enough.css">
<link rel="stylesheet" type="text/yafss" href="web-designer-dreams.yafss">
Received on Friday, 11 December 2009 16:43:38 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 22 May 2012 03:47:12 GMT