RE: iSCSI: Questions regarding text parameter 'Reject' and 'Irrel evant' usage

[Paul Koning <ni1d@arrl.net>]

Excerpt of message (sent 24 April 2002) by John Hufferd:
> 
> My comments were only addressing Boolean and Irrelevant.  Which has been
> discussed in great detail, and what you see is what was agreed.  And these
> functions have also been implemented and have been shown to work in many
> implementation and shown at the plugfests.  The point of last call is to
> bring up items that have not been fully discussed, and resolved.  (You may
> want to review the archives to see what was covered on these topics in the
> past.)
> 
> We need to focus on things that are broken, and editorial fixes.

One way in which specs can be "broken" is that they are ambiguous, so
implementers are prone to do things that don't interoperate.  I think
we have such a case here.

So the spec says you "MAY" send a reject.  What else "MAY" you do?  If
you do one of those other things, what's the originator supposed to
do in response?  It sure does look like something that's broken, and I
don't see the comment being addressed.

      paul