Subject: Re: perhaps time to check our TCP against spec?
To: None <>
From: Jonathan Stone <jonathan@DSG.Stanford.EDU>
List: tech-net
Date: 04/07/1998 18:19:52
>Jonathan, just to be clear, I don't particularly care what you do.  Your
>personal choices don't affect my decisions at all.

 > Further, I am going to tell anyone at Stanford who asks me about which
 > version of Unix to run, that they should NOT choose NetBSD, because
 > Jason Thorpe ( has deliberately made NetBSD's TCP
 > violate historical practice, and that the NetBSD developers have said
 > publicaly that they are not interested in supporting the requirements
 > of instutions which, like Stanford, have large networks, with lots of
 > legacy equipment, designed using the prevalent best-practice rules of
 > the late 1980s and early 1990s.
 > In those exact words.
 > Fair warning.

>I suppose you won't be running BSD/OS, either.

Thanks. I am storing this away and I am going to show it to anyone who
asks me whether NetBSD has *really* stated they aren't interested in
supporting institutions with TCP/IP networks containing legacy hosts.

Personally no, I wont' be running BSDI, and I'd advise anyone who asks
me not to run it on the Stanford net, either. Not if they want
the best possible connectivity with all existing hosts.

>In any case, eventually we'll have Path MTU Discovery enabled by deafault,
>too, so I guess that's one more reason for you to not use NetBSD.  Probably
>means you won't run Solaris or Digital UNIX, either.  Or IRIX, for that

It it's perfectly possible to do  PMTU without in_maxmtu.
The RFC says ``MAY'', not ``MUST''.

Jason, I have pointed out a flaw in your pet idea, and a flaw which
has a real impact.  Because of that, you are resorting to taunts.

Is this an example of what you conisder  ``acting professionally?''