RE: [css3-transforms] Behavior on UAs without 3D support

[Boris Zbarsky:]
> 
> On 5/1/12 5:25 PM, Dirk Schulze wrote:
> > 3) 3D transform functions are treated as invalid if a UA just supports
> 2D. In this case any property settings are rejected if a 3D transform was
> found. Independent if 2D transforms are included in this list as well.
> This gives the author the possibility to provide two different transforms.
> One for UA's with and one for UAs without 3D support:
> 
> This seems like the right approach to me for a UA that doesn't want to do
> 3D.  This is certainly how it would work if they were separate modules and
> the UA just did not support the 3D module.
> 
I agree. It's also consistent with what happens with existing UAs that support
2D but not 3D e.g. IE9. I don't think we'd want authors to have to feature-detect
which kind of feature non-support they're dealing with...