W3C home > Mailing lists > Public > www-style@w3.org > November 2007

Re: Proposal of @ua

From: Andrey Mikhalev <amikhal@abisoft.spb.ru>
Date: Thu, 29 Nov 2007 22:57:34 +0300 (MSK)
To: Andrew Fedoniouk <news@terrainformatica.com>
cc: www-style@w3.org
Message-ID: <Pine.LNX.4.63.0711292225520.11885@master.abisoft.spb.ru>

On Thu, 29 Nov 2007, Andrew Fedoniouk wrote:

>> i mean vendor of some layout engine may support non-standrd properties
>> of another, isn't it? (e.g. to simplify porting/migrating web apps).
>> so, in general, you cannot select 'Mozilla only' rules using single
>> css property check.
> Yes, someone can implement, say, -moz-radius in some other engine. That is 
> why I think that engine specific rules (a.k.a. @ua) are conceptually wrong.
>
> If your design requires exactly -moz-radius style then you will use
> @media screen and supports(-moz-radius) { }
> @media print and supports(-moz-radius) { }
> and that will match all engines that support it. Despite any esoteric reasons 
> why
> engine A would want to implement engine B specific attributes.

imo css feature/property tests and browser/engine tests are things 
targeting two different issues.
first one intended for creation phase where designer choose how page
should look with or w/o modern/vendor css features.
ua-type tests targeting site mantainer who should fix something asap now and 
here, miniminizing posible consequences for other browsers/versions. (yes, 
there are ways to do this not touching css, but...).

> Why you think @else "kills CSS design" and, say,
>
> @media supports(flow,grid) { ... }
> @media not supports(flow,grid) { ... }
>
> is not?

both kills.
backward-compatibility parsing rules, property independency, ...
hence, looks like only this way possible:

@media screen { generic rules }
@media screen and [ua/engine/property/module test] { override rules }

ugly?

>
>
Received on Thursday, 29 November 2007 19:58:00 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:56 GMT