The need for smaller protocol specifications

[Keith Moore]

> Your story, while interesting, is far afield from my point.
> I mentioned several areas which I believe the IESG has
> mistakenly required for inclusion in base protocols, among
> them key distribution when authentication is a part of
> the protocol.
>
> If the IESG wants assurances for a particular technology,
> then those assurances should be in the charter statement.
> They should not be articulated for the first time during
> Last Call.

I suspect we'd agree that having WGs carefully define the scope of the 
problem they are trying to solve, and the applicability of their 
solutions, would help a lot.  IMHO, this should start with the charter 
and continue through a WG "problem statement" phase (what we often call 
"requirements" today) and eventually result in an applicability 
statement that is part of the standard.

Now if we could just stop WGs from claiming vastly more applicability 
than their protocols deserve...