> 3) Looking at the bigger picture, I think this group should look at spoofing as
> a separate group task. It seems to me that there is two sides to spoofing. On
> one side, spoofing is "good" because it addresses the performance deficiency of
> TCP over a satellite link. (and presumable there is no other "standard"
> mechanism that can deliver the performance improvement that spoofing provides)
> On the other side, spoofing is viewed as a "bad" because it breaks the end-to-
> end semantics of TCP and it's long term viability is questionable with
> mechanisms such as IPSEC arriving. It would be my hope that this group can work
> together to converge the two sides together where "spoofing" can evolve to be a
> standard mechanism which enhances TCP performace over a satellite link.
I concur - it would be a worthy effort to constructively address the
topic. If you are going to be at the LA meeting, let's get together and
see how to make this happen.
Firewalls have become firmly embedded into many network architectures;
I've said in the past, these boxes can provide a split connection to
do their job. It's not a major tweak to take advantage of this to
benefit a satellite connection in a (semi-)transparent fashion.
Regards,
Eric
(amazed, but very pleased that something positive might come from
this thread)
This archive was generated by hypermail 2b29 : Mon Feb 14 2000 - 16:14:37 EST