RE: RE: optimizing speed of SMIL

---In [email protected], <ddailey@...> wrote :
 
 > I can’t quite see what you mean by breaking out the two transforms. Do you have an example of that?

Yes, here we are: http://jsfiddle.net/4KWu7/ http://jsfiddle.net/4KWu7/. I think this is easier than having to calculate the required translation animation to counter-balance the scale.
 
 
> In other words, I’m only translating once, so am not sure if your advice would apply.
 
I'm suggesting you don't need an animated translation at all.
 
 
 > While your thoughts are in this mode, if you happen to see any ways of improving performance of any of the examples at http://cs.sru.edu/~ddailey/fading/slides/transitions.html http://cs.sru.edu/~ddailey/fading/slides/transitions.html -- advice would be appreciated.

My suggestion is to try the paint flashing. For filters, you should notice a significant improvement from Firefox 28 (due to bug https://bugzilla.mozilla.org/show_bug.cgi?id=924103 https://bugzilla.mozilla.org/show_bug.cgi?id=924103). But in general, for any browser, if you can find an alternative to tweaking filters it will probably be faster. Also, if you're animating display, you might find visibility is faster.

> There are a couple that don’t work well in Firefox (not sure if is the same as another one I saw, using some now unforgotten search or not) and, of course the one with compound filters applied to the whole window causes some delay.

If you find an example which is clearly slower than others then a bug report is most welcome!


Best regards,

Brian