Re: [css-forms] Some thoughts

> CSS has always been about hinting, rather than absolute specification,
> and the behaviour is further constrained by display medium limitations and
> user physical and virtual (browser dialogue settings) style sheets.
> HTML is not an appropriate technology where designers want guaranteed
> repeatable behaviour.

Of course, that is an _advantage_ of CSS imo. However, since users can 
always overrule designers' decissions they should have the option to 
create their user interface.

I think it doesn't make sense to say that CSS doesn't need a FORM 
specification, when you notice that people like styling forms [1]. I 
think that the W3C should not try to educate people, but try to help 
them by creating useable standards.

Besides that people are already styling forms (with different behavior 
in browsers) some developers create their own widgets with pure JS just 
to get the intended effect. And I must say those look pretty neat.

Why are they creating forms using JS? Because styling them is not 
generalized in browsers. One browser things a form control should always 
be on top of everything else. Another browser doesn't let you style the 
button from an upload field. That is totally inaccessible and imo, 
unacceptable, but that is one a reason not to use the standard HTML 
forms specification.

> In particular, for forms, there is a real expectation, but from simplicity
> of implementation, and usability (users don't want to, and most are
> unable to, learn different users interfaces for each web site), that
> native platform controls should be used.

I hope that "most" is statistically wrong, since a lot of sites are 
using styled forms already.

> For tight control of user interfaces, I would suggest using .NET instead.

I never said I want 'tight control'. I would like a specification which 
addresses FORMS and CSS and specifies which properties apply to elements.

For example:


CSS 2.1 says it is up to the developer what will happen [2]. I think 
that will lead to compatibility problems. (I'm sure, actually.)

[1] <>
[2] Ian Hickson told me.

  Anne van Kesteren

Received on Friday, 9 April 2004 14:37:25 UTC