Xref: utzoo comp.protocols.tcp-ip:10117 comp.dcom.lans:4257 comp.protocols.tcp-ip.ibmpc:2334
Path: utzoo!attcan!uunet!snorkelwacker!spdcc!ima!esegue!johnl
From: johnl@esegue.segue.boston.ma.us (John R. Levine)
Newsgroups: comp.protocols.tcp-ip,comp.dcom.lans,comp.protocols.tcp-ip.ibmpc
Subject: Re: tcp/ip for 386 UNIX
Keywords: UNIX 386 tcp/ip
Message-ID: <1990Feb1.052144.27172@esegue.segue.boston.ma.us>
Date: 1 Feb 90 05:21:44 GMT
References: <5645@ncrcae.Columbia.NCR.COM>
Reply-To: johnl@esegue.segue.boston.ma.us (John R. Levine)
Organization: Segue Software, Cambridge MA
Lines: 20

In article <5645@ncrcae.Columbia.NCR.COM> wingo@ncrcae.Columbia.NCR.COM (Dave Wingo) writes:
> I am looking for a 386 UNIX tcp/ip package and ethernet board combination.

All of the usual 386 Unix vendors ship a tcp/ip package based on Lachman's
port of BSD TCP/IP.  It works as well as any TCP does.  There is NFS on top
of it which Interactive is shipping now and SCO will sometime in the near
future.  There is a little flakiness, e.g. csh filename globbing doesn't
work on remote NFS directories that aren't on System V disks (such as
directories on Suns) but nothing that keeps me from getting my work done.
The TCP/IP interoperates well with other implementations and there are many
386 Unix boxes on the Internet.

As far as the board goes, I'd use the WD8003E which is cheap, reliable, and
has enough buffers to be fast.  Smart Ethernet cards do not improve
performance because the processor on the card tends to be a lot slower than
the 386 that it is attempting to offload.
-- 
John R. Levine, Segue Software, POB 349, Cambridge MA 02238, +1 617 864 9650
johnl@esegue.segue.boston.ma.us, {ima|lotus|spdcc}!esegue!johnl
"Now, we are all jelly doughnuts."