From owner-freebsd-hackers  Thu Oct  9 03:39:55 1997
Return-Path: <owner-freebsd-hackers>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.7/8.8.7) id DAA23080
          for hackers-outgoing; Thu, 9 Oct 1997 03:39:55 -0700 (PDT)
          (envelope-from owner-freebsd-hackers)
Received: from elvis.vnet.net (elvis.vnet.net [166.82.1.5])
          by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id DAA23064
          for <hackers@FreeBSD.ORG>; Thu, 9 Oct 1997 03:39:50 -0700 (PDT)
          (envelope-from rivers@dignus.com)
Received: from ponds.dignus.com (ponds.vnet.net [166.82.177.48])
          by elvis.vnet.net (8.8.5/8.8.4) with ESMTP
	  id GAA12299; Thu, 9 Oct 1997 06:39:43 -0400 (EDT)
Received: from lakes.dignus.com (lakes [10.0.0.3])
	by ponds.dignus.com (8.8.5/8.8.5) with ESMTP id GAA08871;
	Thu, 9 Oct 1997 06:54:48 -0400 (EDT)
Received: (from rivers@localhost) by lakes.dignus.com (8.8.5/8.6.9) id GAA06418; Thu, 9 Oct 1997 06:45:32 -0400 (EDT)
Date: Thu, 9 Oct 1997 06:45:32 -0400 (EDT)
From: Thomas David Rivers <rivers@dignus.com>
Message-Id: <199710091045.GAA06418@lakes.dignus.com>
To: fenner@parc.xerox.com, jonny@coppe.ufrj.br
Subject: Re: TCP problem
Cc: hackers@FreeBSD.ORG
Sender: owner-freebsd-hackers@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

> 
> Joao Carlos Mendes Luis <jonny@coppe.ufrj.br> wrote:
> >  I have an intermitent TCP problem between a FreeBSD 2.2-STABLE and a
> >2.0.27 Linux.  It's happening right now, let me show an example:
> >22:49:35.576825 146.164.5.200.2038 > 146.164.53.91.19: . ack 1 win 164 (DF) [t
> >  os 0x10] (ttl 64, id 38632)
> 
> This packet ("win 164") is the exact symptom of the problem that was fixed
> in rev 1.27 / 1.21.2.3 of tcp_output.c .  Funny that everyone's coming
> across this one all at once.
> 
>   Bill
> 
 
 Ah... then you've not enjoyed my "lecture" on my theory of "bug time."

 The basic premise is based on empirical experience; not on any theoretic
foundation... but, it seems that, sometimes major, bugs will lie in the
code; domant - for many years.  Then, suddenly, a confluence of events
all around the world will cause the bugs to manifest themselves.  [I call
such bugs "locusts".]

 Usually, an addenda to the theory indicates that there is no way the
software could have been resounably exercised (by a user or testing group)
without finding this bug; which means:

	1) There exists some force which can go back in time and
	   inject bugs that weren't previously there...  So, for
	   some window of time after that point - the software works
	   then stops working.

	2) The software hasn't been exercised (tested) as much as
	   one would expect (hope.)

	:-) :-)

	- Dave Rivers -