At 02:35 AM 12/4/00 -0800, pmoore at peerless.com wrote:
>3. conneg
>>We have had some exchanges about the strengths and weaknesses of conneg.
>Nobody
>has come up with a better solution - anybody with any ideas is welcome to
>speak
>up during this slot
I'm not sure if some back-channel discussions have filtered back to this
group, but the relatively recently published RFC 2938 was intended to
alleviate some of the difficulties that have been noted.
In particular, selectable feature-group expressions may be replaced by a
composite identifier, so that the resulting feature expression is
simpler. This may be particularly useful for removing inner disjunctions
from the resulting expression.
#g
------------
Graham Klyne
(GK at ACM.ORG)