Re: proposal: @supports and :supports

Testing for feature support _does_ work in practice.
Where? With JavaScript and the DOM. Although there are
rare occasions where browsers have lied about what
they support, for the most part it works very well.

Think about events, which are very important. The W3C
specifies addEventListener, but IE uses attachEvent.
Whithout the ability to test for support, using events
woud be much more difficult.

Testing for feature support is working _right now_.
It's used by the utmost professionals. History has
proven that although testing for support isn't
perfect, it definately _is_ very beneficial.

--- Ben Ward <ben@ben-ward.co.uk> wrote:

> Off the top of my head, it fell down on a number of
> issues, one of which 
> centred around the requirement for user agents to
> implement it correctly 
> and be honest about which CSS features they support.
> Furthermore, the 
> concept of supporting a particular CSS feature does
> not indicate whether 
> the feature is supported _accurately_, without bugs.
> As such, the value 
> of features like this has been shown to evaporate
> quite quickly and it 
> all comes out looking a bit too Utopian to work in
> practice.


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com

Received on Thursday, 25 May 2006 01:03:10 UTC