Re: [CSS21] Lack of version control for content

> If the CSS WG does not consider that validation of CSS stylesheets is
> appropriate, please explain that to the CSS validator folks. Currently,
> the W3C CSS validator seems to assume CSS 2.0, so it complains about
> orange and fails to complain about properties in CSS2 and not in CSS
> 2.1, for example.
> 
> Similarly, if someone is using CSS MP, the validator woudl fail to warn
> about using properties outside that profile.

Surely those are flaws in the validator, rather than in the spec?

They could be rectified in the future with an improved validator.
Through these recent discussions individuals who work on the validator
have, I believe, commented that a flexible 'profile' system for the
validator, separate to any kind of CSS 'versioning', would be more
beneficial to page authors.

Such a profile system sounds excellent, and I'll be a happy many if
the validator volunteers can find the time to create such a tool.

> Validation is useful.
> Adding an @-rule would be an easy way to indicate to which specification
> a given style-sheet conforms.

Yes, validation is useful, but I fail to see how confining someone to
a specific CSS revision helps authors. I create a CSS2.1 stylesheet
but want to use CSS3 opacity and multiple backgrounds, what do I do?
What does a user agent do when it encounters a higher version number
than it understands? None of these things are handled intuitively by
such hard versioning. CSS works much better in its flexible state.

How would an @rule cope with CSS3's module based specification?


Seriously, adding version information rubs against the grain of CSS
and is a validator issue, not a CSS spec issue. It's not just against
the way in which the spec appears to be prepared and written
(referencing CSS3 work, for example), but also grates against the way
in which CSS is used in the real world by authors.

Ben
http://ben-ward.co.uk

Received on Tuesday, 30 August 2005 08:20:23 UTC