Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!bbn!uwmcsd1!ig!jade!ucbvax!ftp.UUCP!jbvb
From: jbvb@ftp.UUCP (James Van Bokkelen)
Newsgroups: comp.protocols.tcp-ip.ibmpc
Subject: Re: 3COM 3c503 (Etherlink II) -- info needed.
Message-ID: <8711190649.AA01789@spdcc.COM>
Date: Thu, 19-Nov-87 01:35:44 EST
Article-I.D.: spdcc.8711190649.AA01789
Posted: Thu Nov 19 01:35:44 1987
Date-Received: Sat, 21-Nov-87 15:34:06 EST
References: <8711190246.AA23866@umd5.UMD.EDU>
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The ARPA Internet
Lines: 31

I have a driver I wrote myself in beta test.  I am not deeply impressed with
the "3L" interface:

	1. It has a lot of bells and whistles I didn't need.

	2. If you want to use it out of the box, you have to live with the
	fact that it is designed for a TSR LAN program, and tries to parse
	command-line arguments on startup.

	3. It is all 3Com copyrighted code, even though they supply source.
	If permission to redistribute it is available, I haven't seen it.

	4. It doesn't buy you hardware-independence at the .EXE level - you
	have to link in a specific driver for one board or another.  We are
	eagerly awaiting the "vector" interface, but the spec hasn't appeared,
	even under non-disclosure.

The 8390 chip has some bugs, which show up in some other cards at high data
rates, or only at Starlan clock speeds.  Possibly because of the gate array,
these bugs show up full-blown during normal use of the 3C503.  My advice to
the punter is: Reset it thoroughly every time it does something suspicious,
and start with a production card (the sticker on it had better not say
"IE6PTYP...").

I have actually seen a programming manual for the 3C501, but it is a true
fossil, having been given to MIT back when the original PCIP was being
developed.  I believe that current versions are not at all easy to obtain,
nor have I seen a "3L" for the Etherlink.

James B. VanBokkelen
FTP Software Inc.