Claude L Bullard wrote:
> don't normatively associate any spec with any
> other spec unless you have to.
It is a standard rule of specification efforts that the
likelihood of useful output in a timely fashion is inversely
proportional to the number of dependencies.
I support all who have argued that the TC should stay aware
from user interface issues.
In any case, before you can deal with UI issues, you need to
have a solid data model first. Thus, if folk want to move in this
direction, the first challenge would be to define schemas for the data
-- independent of presentation. When doing system architecture, be
sure to clearly distinguish between "form, function, and data
management." (i.e. the "three layers"...)
Personally, I think that the number 1 priority for the TC
should be to develop a more useful CAP. (i.e. something other than
that which was just voted on...)
bob wyman
