- From: Dr. Olaf Hoffmann <Dr.O.Hoffmann@gmx.de>
- Date: Thu, 1 Nov 2007 12:11:09 +0200
- To: www-svg@w3.org
> the current Opera nightlies animate the attachment, though this is > stated to be a regression. >What is the reason the WG chose not to support the matrix value? > >it's not immediately clear to me how else this behaviour might be >implemented... Surprise, surprise! However this is a very useful improvement, but should not happen for a defined version like 1.0, 1.1 or tiny 1.2 ;o) I already did some simulations using an animation of xlink:href to simulate a frame-based animation and there is no technical problem to animate matrix. The 'frame-based' method only blows up the source code with superfluous constructions. Maybe it is only excluded in the recommendations for historical reasons? Maybe the idea was, that it is a too complex construction to be animated? (just guessing) Of course especially in the filter chapter there are more complex attribute constructions animatable as the type matrix of transform. Maybe the matrix type should be animatable in SVG 1.2 full? I would like to use it. Something has to be fixed before - there are several problems with implementations of animations only using the 'to' attribute and a minor with 'by' and some general inconsistencies with additive animations for animateTransform in general. But if this is clarified or aligned to a useful behaviour, there is no really problem left with the type matrix, I can see. Because authors are not allowed to animate matrix in the current profiles, it is not really a problem, if it is possible with Opera anyway, because in a correct document this will never appear and there is no chance for Opera to animate this because it simply does not exist in a correct document ;o) And why to be fussy with the interpretation of faulty documents? ;o)
Received on Thursday, 1 November 2007 10:17:53 UTC