W3C home > Mailing lists > Public > www-style@w3.org > April 2010

Re: Are CSS animations a done deal?

From: Tab Atkins Jr. <jackalmage@gmail.com>
Date: Mon, 5 Apr 2010 08:55:37 -0700
Message-ID: <m2vdd0fbad1004050855p5a7ce004i9aa37b42c01ff565@mail.gmail.com>
To: Lars Gunther <gunther@keryx.se>
Cc: www-style list <www-style@w3.org>
On Mon, Apr 5, 2010 at 8:33 AM, Lars Gunther <gunther@keryx.se> wrote:
> 2010-04-05 12:32, sam skrev:
>> Being able to specify styles as declaration blocks in CSS as a class and
>> reference/apply them in JS via className is good (but maybe class name
>> is not the right mechanism to expose declaration blocks to JS enviroment).
> Let's consider what developers are going to want to do:
> 1. Apply animations to any kind of event. (A desire augmented by the fact
> that there is no hover on the iDevices...)

They can do that just as easily with CSS Transitions & Animations
(hereafter known as T&A) as with a javascript-based one.

> 2. Change animation properties, like individual rules in a key frame.
> Not possible at all with the current proposal.

This will be defined in due course by the CSSOM.  Significant work is
going on there, both to actually make it *work* and to make it easy.

> 3. Use animations for effects that are not pure eye candy, and thus carry
> some kind of information. Where is the fallback or better yet the built in
> accessibility for this?

Can you give an example of such?  I can't think any place where
Transitions might be informative rather than pretty.  I'm not sure
about Animations yet.  You can certainly transmit *some* information
using them, just as you can with any CSS (for example, you might color
the border on a table-cell to indicate the value of some variable).
Do you have any examples of cases that go beyond this?

> Declaring animation properties in JSON and applying the animations in
> DOM-events addresses these three concerns:
> 1. Although no new functionality is added this way, the solution is cleaner.
> At least, IMO.

In your opinion, certainly.  I think the opposite, so we're even.  ^_^

> 2. It makes this possible. Currently it is not.

It makes what possible?

> 3. It integrates better with ARIA and other accessibility techniques.
> That last point needs some careful evaluation, and indeed the whole issue of
> animation needs some accessibility evaluation, but it seems to be an easier
> path to me upon first inspection.

Again, I'd like some more explanation of what you mean by this.
Assuming there are cases where animations carry semantic value, I'm
not sure how putting them in JS is better than CSS.  They appear to be
equivalent in terms of accessibility.  After you've formulated some
cases where animations carry actual information (beyond the natural
sort of information that CSS can carry anyway), could you explain
precisely how moving such an animation to a JSON format improves

Received on Monday, 5 April 2010 15:56:27 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:38:33 UTC