W3C home > Mailing lists > Public > www-style@w3.org > November 2003

Re: CSS version selectors?

From: David Woolley <david@djwhome.demon.co.uk>
Date: Sat, 1 Nov 2003 10:09:23 +0000 (GMT)
Message-Id: <200311011009.hA1A9No06039@djwhome.demon.co.uk>
To: www-style@w3.org

> Of course, this has the option of being abused because a browser-maker 
> could implement the version selector without fully implementing the rest 

It would also be abused because they would delude themselves into believing
that they did fully conform and they would take the most liberal possible
interpretation of the conformance requirements.  Also remember that CSS
provides hints and implementations don't have to accept those hints, and
hardware may sometimes prevent their being implemented.

Currently no browsers fully conform.  Even if you included the media
type in this (and therefore required that it only be used within an 
@media section), I don't think that you would find any browser that could
match one of these selectors currently.

In general, I believe that this sort of adaptability needs to be done by
specific feature tests, never by browser sniffing and probably not by
an overall version compliance claim.  The feature tests need to go 
beyond whether the browser supports the feature, they must also take
into account whether it will be effective in the current context; users
are allowed to override authors and conditionals need to account for that.
(E.g. IE 5.5 provides fairly user friendly ways of selectively disabling
any comibination of colours, font-faces and font sizes, and a techie 
interface for providing a general overriding style sheet.)  In some cases,
display hardware may impose constraints as well.

Even with feature tests, one is at the mercy of a vendor's creative 
interpretation of compliance.

My own feeling, is that the right approach is to provide a way of
bracketing a set of rules to indicate that they should all be honoured or
all rejected.  The author then puts a fall back set first, followed by
groups in ascending order of preference.
Received on Saturday, 1 November 2003 05:19:05 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:24 GMT