RE: [cgmo-webcgm] about transform [was RE: [cgmo-webcgm] Groups - Proposed WebCGM 2.1 ...]

[Lofton Henderson <lofton@rockynet.com>]

At 12:21 PM 9/4/2007 -0400, Weidenbrueck, Dieter wrote:
>[...]
>Re 2)
>Agreed, and it makes more sense.
>This requires a different approach than style properties, because
>setting one of those replaces any previous entry. In the transform case,
>we need to concatenate.

Yes, we would have to specify, if we were to endorse such a set of 
transform components, that they accumulate (concatenate), unlike the other 
Style Properties.

-Lofton.


>-----Original Message-----
>From: Lofton Henderson [mailto:lofton@rockynet.com]
>Sent: Dienstag, 4. September 2007 18:08
>To: Weidenbrueck, Dieter; Bezaire, Benoit;
>cgmo-webcgm@lists.oasis-open.org
>Subject: RE: [cgmo-webcgm] about transform [was RE: [cgmo-webcgm] Groups
>- Proposed WebCGM 2.1 ...]
>
>Dieter,
>
>Thanks for the detailed contribution.  I'll look at all of it more
>carefully, but one small technical point catches my attention now...
>
>At 11:44 AM 9/4/2007 -0400, Weidenbrueck, Dieter wrote:
> >Lofton,,
> >
> >There are several issues with this:
> >
> >- user experience
> >         We wonder how many users are capable of building transform
> >matrices by concatenating them using EcmaScript. This requires quite a
> >level   of mathematical skills.
>
>Indeed.  But two observations:
>
>1.) in scenarios like Molly/Larry, their tool would presumably provide a
>simple parameter interface of some sort to their customers, and hide the
>dirty work of generating the transforms.
>
>2.) SVG is interesting in providing, in addition to 3x3 Matrix,
>individual transform components for Scale, Rotate, Translate.  (Of
>course, that presents a whole different set of traps for the careless
>users -- transform components do not commute, etc.)
>
>-Lofton.