From: Russ Allbery (rra@stanford.edu)
Date: Sat Jan 06 2001 - 21:33:46 CST
Charles Lindsey <chl@clw.cs.man.ac.uk> writes:
> Russ Allbery <rra@stanford.edu> writes:
>> Yeah, this really is something that, if in the standard, makes me feel
>> like I should raise it at Last Call as an inappropriate use of RFC 2119
>> terminology. It certainly is outside our scope.
> This is clearly a borderline case.
I honestly don't agree. I think there are tons of borderline cases
already in the standard that I've not tried to pursue in the interests of
speeding things along, and this one is clearly over the line. But
apparently that opinion isn't universally shared, and I really don't want
to bog the group down into a big argument about it.
I have a fairly strict concept of what a standard should be. I realize
that other people have a much broader concept, and want to include various
hints, suggestions, best practices, and other sorts of information for
implementors. The part of my brain that does standards work is the same
part that does mathematical proofs.
Anyway, I expect the whole "ought" thing to come up again later at Last
Call, and I expect that it will get resolved then, so we may as well move
on for the time being. No point in having the argument twice.
-- Russ Allbery (rra@stanford.edu) <http://www.eyrie.org/~eagle/>