Re: Protocol: control specifications.

[Hallvard B Furuseth <h.b.furuseth@usit.uio.no>]

Kurt, I'm suddenly wondering if I've misunderstood you completely,
and if so I don't know where.  So I need a bit of clarification
before replying to your latest message:

You are saying, I think, that current practice is for most servers
to ignore the criticality protocol field when the criticality is
given by the control spec.  Yet you suggest that control specs
should no longer say what the criticality _is_, they should merely
give guidance to what the field should be _set_ to.  Thus, there
will no longer be a criticality in the control spec which servers
can follow, so they all MUST change from current practice to always
obeying the protocol field.

Yet one of your arguments against my suggestion is that it changes
current practice.  My suggested change is merely a SHOULD.  Servers
may ignore it and keep doing what they do now.  Or if you disagree
about that, I may have made a mistake and should modify my
suggestion, but we can get back to that later.

-- 
Hallvard