[3471] in linux-net channel archive
Re: shaper or whatever
daemon@ATHENA.MIT.EDU (Alan Cox)
Mon Jun 24 20:24:36 1996
From: Alan Cox <alan@cymru.net>
To: dennis@etinc.com (Dennis)
Date: Mon, 24 Jun 1996 18:50:56 +0100 (BST)
Cc: alan@cymru.net, linux-net@vger.rutgers.edu
In-Reply-To: <199606241733.NAA16949@etinc.com> from "Dennis" at Jun 24, 96 01:33:54 pm
> Perhaps the issue is that I'm addressing what the marketplace wants,
> which is physical simulation, and you're addressing some other issue
> that I can't quite figure out. Different goals, different implementations.
> Its not really that surprising. The proof is in the pudding (as someone
> once said), it's clean and works as intended, which is what its all
> about.
Im addressing what people want which is to give people the bandwidth they
want without causing excess load elsewhere on the network. Im not sure
what you are doing but it doesnt seem to involve the latter rule. For
reference Im holding a 64K over ethernet at +- 3% on the test rig and
causing no measurable excess retransmissions.
Alan
