Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!gatech!hao!noao!mcdsun!sunburn!dover!leivian
From: leivian@dover.uucp (Bob Leivian)
Newsgroups: comp.dcom.lans
Subject: Re: Ethernet configuration
Message-ID: <6@dover.uucp>
Date: Fri, 8-May-87 13:01:25 EDT
Article-I.D.: dover.6
Posted: Fri May  8 13:01:25 1987
Date-Received: Sun, 10-May-87 03:32:35 EDT
References: <331@umbc3.UMD.EDU>
Reply-To: leivian@dover.UUCP (Bob Leivian)
Organization: Motorola CAD Mesa, AZ {dover}
Lines: 28
Keywords: Ethernet DACU SEICOR

> spartacus vs DACU
  we have a large TCP/IP network with IBM and sun/apollo/pc workstations
  we evaluated the two boxes for IBM a year or so back and spartacus was
  a clear win, the DACU uses a PC to control a funny dr11 kludgey (will
  someone please establish the correct spelling for that once and for
  all) interface that was 3 or four times as slow as sparatcus which
  had a channel attached regualar IBM device.  The only software
  avail at the time was U. of Wisc for the DACU, while it was al right
  it did not match the commercial code from spartacus.

  while we have had some problems with other tcp/ip interfaces because
  spartacus is written in IBM ASM and is not state machine driven like
  the BBN and BERK. versions that most are ported from it matches
  pretty closely.  some other implementations that are written from
  scratch from the spec have problems talking (the appletalk/tcp gateway
  'fer instance sends the initial DO/DON'T sequence with the ack
  of the SYNC where as most send a separate ACK)

  > transfer rates
    we are seeing rates (binary) of about 50,000 bytes/sec to and from
    IBM using spartacus and some of our faster workstations (sun's).


    ---- these are my opinions and do not necessarily reflect Motorola
      
-- 
Bob Leivian         Motorola, Dover Shores (CAD Support)  602-994-6778
 ...{mcdsun, sun}!sunburn!dover!leivian