Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!mit-eddie!genrad!decvax!decwrl!ucbvax!A.ISI.EDU!CERF
From: CERF@A.ISI.EDU
Newsgroups: comp.protocols.tcp-ip
Subject: Re: TCP performance
Message-ID: <[A.ISI.EDU].4-Oct-87.06:37:44.CERF>
Date: Sun, 4-Oct-87 06:37:00 EDT
Article-I.D.: <[A.ISI.EDU].4-Oct-87.06:37:44.CERF>
Posted: Sun Oct  4 06:37:00 1987
Date-Received: Wed, 7-Oct-87 05:49:28 EDT
References: <art@acc.arpa>
Sender: daemon@ucbvax.BERKELEY.EDU
Distribution: world
Organization: The ARPA Internet
Lines: 12

Art,

indeed, delay is the ultimate arbiter - but your figures left out
the lower delay paths for smaller nets (local ones).  I think we
ll agree, though, that window-based schemes suffer when long delays
are involved and high speeds. The source must be tolerant of 
very long relative delays before receiving acknowledgement so
hat pacing rather than buffer acks has to be the flow control
method because buffer acks come at too long delay to be an effective
low control - the loop time is too long.

nt