Subject: Re: Socket buffer accounting and TCP
To: None <firstname.lastname@example.org>
From: Bill Sommerfeld <email@example.com>
Date: 09/03/1998 09:40:27
Rather than speculating about the virtues of different compression
strategies, I'd like to suggest a meta-approach -- get some traces of
*actual* socket queue mbuf pileups as food for analysis. We can then
simulate different compression heuristics and see which ones provide
the most gain for the least cost.
I may spend a little time on that this morning....