RE: comments on draft-ietf-cain-request-routing-req-00.txt

["Mark Day" <markday@cisco.com>]

> Or is this all now outside of the WG charter no longer
> requiring the protocol specification and, hence, protocol
> design considerations?

I think what we want to do is to recast your comments into terms that can be
captured in a requirements document.

If you can describe your issue in terms that apply to the overall problems
of request-routing, distribution, and/or accounting, then it's within our
charter.  I believe that protocol design considerations are important parts
of the requirements -- that is, we should require that protocols have
known-good features and also require that they not have known-bad features.

I don't have a firm opinion about whether your particular concerns belong in
the requirements or not, but protocol design considerations per se are OK
for discussion.  That's especially true when (as in this case) the
discussion is about concrete (mis)features of particular proposals.  I'm
less supportive of feel-good requirements like "the protocol should be
simple" or "the protocol should be efficient."

--Mark