- From: Dr. Olaf Hoffmann <Dr.O.Hoffmann@gmx.de>
- Date: Wed, 9 Jan 2008 12:52:03 +0100
- To: www-svg@w3.org
If color itself is animated as in animate-elem-85-t it will not be sufficient to pass the test, if the currentColor is only determined at the beginning of another animation. It has to be determined any time, the value is needed, if color is changed somehow. I think, this behaviour can already be determined from the property name, it is currentColor and not pastColor or futureColor or something like this. For example in the bottom subtest of animate-elem-85-t the resulting color changes in a quadratic dependence, because both fill and color are animated and currentColor is the to-value for fill. This is somehow similar to the behaviour of to-animations with underlying animations, obviously not really easy to implement, but for any example it is precisely testable, whether this is implemented correctly or not. And currently, I think it is risky for authors to rely on such more complex animations unfortunately.
Received on Wednesday, 9 January 2008 12:08:45 UTC