RE: [css3-2d-transforms][css3-3d-transforms] Relationship of CSSMatrix interface definitions from Sylvain Galineau on 2010-09-08 (www-style@w3.org from September 2010)

From: Sylvain Galineau <sylvaing@microsoft.com>
Date: Wed, 8 Sep 2010 21:31:32 +0000
To: Chris Marrin <cmarrin@apple.com>
CC: Dean Jackson <dino@apple.com>, "www-style@w3.org list" <www-style@w3.org>
Message-ID: <045A765940533D4CA4933A4A7E32597E27CACD01@TK5EX14MBXC120.redmond.corp.microsoft.>

> From: Chris Marrin [mailto:cmarrin@apple.com]


> >> That looks much more clear, to me at least. Is it legal to redefine
> the
> >> translate, scale and rotate functions?
> >
> > Legal in what sense ? It certainly makes sense to add the optional z
> parameter
> > to the 3D interface and it seems cumbersome to have two versions of
> each
> > method.
> 
> I mean is it valid IDL. If so, then no problem.

Oh, got it. You're right. These are not strictly speaking [Supplemental]
methods; they have to replace the existing translate(), scale() and rotate()
or you can't tell which one applies when there are only two arguments.

This also implies both sets of matrix properties are updated by those methods.
It's one of those things worth calling out for completeness.

Received on Wednesday, 8 September 2010 21:32:11 UTC