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

Re: [CSS] Extending media queries

From: Joshua Cranmer <Pidgeot18@verizon.net>
Date: Tue, 17 Feb 2009 13:11:02 -0500
To: CSS 3 W3C Group <www-style@w3.org>
Message-id: <499AFDB6.8050102@verizon.net>
François REMY wrote:
> _Here's a bit more about how I imagine my proposition :_
> *@media (implements: position) { ... } *
> >>> True if the browser has a CSS property called 'position', even if 
> this property is not standardized or W3C non-conformant.
As previously pointed out, how would this help when you discover that 
layout engines manage to misread many properties? A common example is 
IE's lack of the 'inherit' keyword; another one I ran into was that it 
(at least IE 6) miscalculated left. The cases were a browser 
misimplements a property is probably as important to an author than 
where it doesn't implement the feature at all.
> *@media (implements: position 2.1 strict) { ... } *
> >>> True if the browser fully (whitout any known bug) supports the 
> 'position' property as the W3C defines it. If the browser is build 
> before the standard is completely OK, then it should return False, 
> even if he thinks the standard will not change and that all is ok.
Define "without any known bug." Should a browser say yes, even if the 
known bug in question happens when you jumble together floats, relative, 
absolute positioning with inline tables and boxes with 10-pixel widths 
in crazy corner cases? Who gets to be the arbiter? The person who first 
implemented the feature probably expects it to work in all cases, but it 
might subsequently be discovered that it misses big cases...
> *@media (implements: position 3 drafts) { ... } *
> >>> True if the browser mostly supports the 'position' property as the 
> W3C defines it in a draft posted in the CSS 2.1 section. If the 
> browser is build before the standard is completely OK, then it may 
> return True.
I don't see the utility of this at all. Drafts are not necessarily 
stable--the first draft of CSS 3 Über Feature probably isn't defining 
the same things as the last call for said module. So your proposal 
wouldn't be able to distinguish between the layout engine implementing 
the first feature set and another implementing the second.

In short, as with all the previous @has-feature-esque proposals, I 
foresee a lot of problems with this proposal.

Beware of bugs in the above code; I have only proved it correct, not tried it. -- Donald E. Knuth
Received on Tuesday, 17 February 2009 18:12:02 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:38:24 UTC