SpecGL CR criteria (was: Draft minutes of Jan 5, 2005 telecon)

Le jeudi 06 janvier 2005 à 17:43 -0700, Lofton Henderson a écrit :
> At 09:23 AM 1/6/2005 +0100, Dominique Hazaël-Massieux wrote:
> >[...]
> > > KD: we have been told during CR that we must have two complete 
> > implementations
> >
> >should read "We had been told at our previous CR transition call that we
> >must show 2 complete implementations"
> 
> This confuses me.

For what it's worth, I'm only reporting what I heard Karl say.

>   The criterion that is used by SVG and almost everyone 
> else is:  show two implementations of each feature.

... and I think this is what was indeed required from us, not the
former.

> That is very different from two complete implementations.  Are we really 
> being held to "two complete implementations"?  If so, why is that being 
> applied to QAWG, but to no other WGs?

From rom what I remember, our CR exit criteria had this additional point
that our implementations should show that groups were able to use SpecGL
on their own, i.e. apply the principles in it based on their reading of
the spec rather than simply by their own wisdom...

But Karl may remember differently...

Dom
-- 
Dominique Hazaël-Massieux - http://www.w3.org/People/Dom/
W3C/ERCIM
mailto:dom@w3.org