-----Original Message-----
From: Graham Klyne [mailto:GK-lists(_at_)ninebynine(_dot_)org]
Sent: Thursday, February 21, 2002 3:40 PM
If every hare-brained idea for a new header gets a permanently
allocated name to satisfy this condition, then the interesting
parts of namespace are potentially liable to to be committed.
I think it's helpful if a degree of community consensus is required
to obtain a unique, permanent allocation.
The interesting parts of the namespace only get consumed if the URN of
the field is based only on the name of the field. Were the URN for
fields to be a stricly increasing number (like RFC numbers) or the field
name plus a strictly increasing number (like ID names), the namespace
for *fields* (as opposed to URNs) would be unaffected by hare-brained
schemes.
-- jeff