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

Re: [animations] fill modes, revised

From: Dr. Olaf Hoffmann <Dr.O.Hoffmann@gmx.de>
Date: Wed, 3 Mar 2010 10:57:58 +0100
To: www-style@w3.org
Message-Id: <201003031057.58887.Dr.O.Hoffmann@gmx.de>
Simon Fraser:
>Note that SMIL doesn't do backwards fill (before the animation starts) and 
>uses different keywords. I think the ones below make more sense.
>animation-fill-mode:  none | forwards | backwards | both;

I think, a backwards fill does not make much sense in SMIL animation
due to the advanced sandwich model and the priorities of different 
animations for the same attribute or properties.
And it is simple to use just another lower priority animation or a static
underlying value to get this effect. For video however SMIL (and 
SVG tiny 1.2) has such an effect, just because it is not trivial to get this
effect to provide the first frame before active duration as a preview,
it is just another attribute.

Some forward-fill for CSS-animation would be indeed pretty
helpful for authors and because CSS-animation is much simpler
(and has therefore less options for authors), it should be simpler
to get the intended effect implemented for CSS-animations than
for the SMIL with its advanced model with multiple begin and
end times and corrections of active durations (active duration can
be longer than simple duration or repeated simple duration), what
is typically pretty useful as well (including event-based begin and end).

To use the same keywords as in SMIL/SVG would have the
advantage of better compatibility and that authors do not have to
learn two vocabularies for the same effect, maybe something 
for the CSS/SVG FX taskforce to align the CSS vocabulary to that of

Received on Wednesday, 3 March 2010 10:00:33 UTC

This archive was generated by hypermail 2.4.0 : Friday, 25 March 2022 10:07:43 UTC