Date: Wed, 19 Sep 2001 09:46:21 -0500
From: "David Fischer" <>
Very good. I agree with most of it.
Thank you.
One comment about check-sums. We already have an transmission error-catching
mechanism called NRR.
Yes, but it has several drawbacks as an error-catching mechanism.
First, it's optional. Second, the error isn't detected until after
the receiver might have processed the message.
On the whole, I think this is very good. The point is that there are some
scenarios which would require a retry. But I prefer to phrase the question
differently -- why would an IM *ever* stop a retry from the end? It is not the
job of an IM to tell the ends what they may or may not send. Since it is an
easy thing to differentiate between an IM retry and an end retry, its not "why
would we" but rather "why wouldn't we"?
Of course an IM should never stop an end-to-end retry; of course if
the protocol allows for end-to-end retry, there would be an easy and
well-defined way to differentiate. The proposed "retry count" is one
way to do it.
