RE: [css3-transform] definition of skewing

Hi Chris,

SVG probably doesn't do transitions and animations like CSS, so most likely it wasn't a problem there. 

It's probably OK not to add rotate(x, ) for now because there is a workaround by using multiple discrete matrices.
This is not too hard if you're generating the CSS programmatically but for a regular user, it will be hard to calculate and it also creates bigger CSS.

Like I said, this is a common transformation in all Adobe apps and probably other applications as well so it seems that it should be part of the spec.


> -----Original Message-----
> From: Chris Marrin []
> Sent: Wednesday, February 02, 2011 5:48 PM
> To: Rik Cabanier
> Cc: Simon Fraser; Alan Gresley;
> Subject: Re: [css3-transform] definition of skewing
> On Jan 25, 2011, at 10:55 AM, Rik Cabanier wrote:
> > Hi Chris,
> >
> > There are 3 facts in this discussion:
> > 1. with the current state of the spec, you can't create this type of 2D
> animation without going to a 3D space. Even then, the result looks different.
> > 2. having skewx/y as opposed to rotate(x,y) as part of an animation can
> introduce unintended behavior where the graphic will either become very
> large or go to 0.
> > 3. This is a transformation that is used quite often by animators since it's
> easy to find examples on the web.
> You omitted one fact: SVG has a set of transform functions that does not
> include your notion of rotate(x,y). I think it would be a mistake to blaze a trail
> here unless we were absolutely certain it was the right set of new
> functionality. In the future we might want take a crack at unifying the Matrix
> functionality between the two and that might be a good time to refine the
> API. But in my opinion we shouldn't add this function in the first release.
> -----
> ~Chris

Received on Friday, 4 February 2011 03:31:36 UTC