- From: L. David Baron <dbaron@dbaron.org>
- Date: Tue, 27 Sep 2011 18:14:00 -0700
- To: Nicholas Shanks <contact@nickshanks.com>
- Cc: www-style@w3.org, Ian Hickson <ian@hixie.ch>
On Monday 2011-09-12 17:00 +0100, Nicholas Shanks wrote: > I propose a new, optional, condition keyword: 'all', to be used thus: > > @supports all { > .box { > display: inline-block; > box-spacing: even; > } > } > > Which is equivalent to the following: > > @supports { > .box { > display: inline-block; > box-spacing: even; > } > } I'm more comfortable with the explicit "@supports all" variant than I am with assuming that @supports with no argument is equivalent to it. However, I worry that by not asking authors to explicitly declare what they require, we might be tempting them to unnecessarily create more dependencies than they need. This would increase the chance that authors will accidentally write browser tests rather than feature tests (by limiting things to the set of browsers they test on and accidentally excluding others) It also fits rather awkwardly with CSS's existing mechanisms for ignoring things that are unsupported. CSS has been careful to define the scope of what an error invalidates (e.g., declaration, rule, media query expression), and in general define things thus invalidated to act as though they were not present. When such things go inside @supports all, this both breaks the mental model that I hope authors develop of CSS's error handling and the model that implementations use to implement it. -David -- 𝄞 L. David Baron http://dbaron.org/ 𝄂 𝄢 Mozilla http://www.mozilla.org/ 𝄂
Received on Wednesday, 28 September 2011 01:14:49 UTC