[comp.protocols.tcp-ip.ibmpc] PC/IP ScoreCard - Revision 3

HANK@BARILVM.BITNET (Hank Nussbacher) (03/10/88)

                         The Pc/Ip Scorecard
                        revision 3:  03/10/88
                        ---------------------

    We   are   the   process   of   trying   to   decide which Pc/Ip
implementation  would  be  the   best.   After  asking  around   the
overwhelming reply  I have  received has  been, "If  you ever find a
comparison study, I would love to  see it too".  That is why  I have
decided to create the Pc/Ip Scorecard.

    This scorecard will  be like a  PC Magazine analysis  of of hard
disks or printers.  But  I need help in  filling in the boxes.   So,
here is what the scorecard looks like.  Please send me your  replies
and  I  will  integrate  all  answers  and  comments and publish the
finalized scorecard in the weeks to come.

    All Pc/Ip implementations support IP, TCP, FTP and Telnet.   The
question then  is to  divide other  IP protocols  into categories of
"must have" to "nice to have but not necessary".

|Revision 3: Any implementation that has not had any sort of comment
|            made about it over the past 3 weeks has been removed.
|            There are still plenty of gaps in the ScoreCard.  Please
|            help out.


    This first table is called "Must Have".

Vendor      TFTP  POP ICMP SMTP VT-  3270 FTP   ARP  UDP max  cost
                                 100                     FTP  ($)
-----------+----+----+----+----+----+----+----+----+----+----+----+
Beame      | Yes| No | Yes| No |    | No |    | Yes| Yes|    |    |
CMU        | Yes| No | Yes| No |    | No | No | Yes| Yes|    |   0|
Excelan    | No | No | Yes| No | Yes| No | Yes| Yes| Yes| 42k| 250|
FTP        | Yes| No | Yes| Yes| Yes| Yes| Yes| Yes| Yes|160k| 400|
IBM        | Yes| Yes| Yes| Yes|    | Yes|    | Yes| Yes|    |    |
KA9Q       | No | No | Yes| Yes| No | No | Yes| Yes| Yes|    |   0|
MIT        | Yes|    | Yes|    |    |    | No | Yes| Yes|    |  50|
NCSA       | No | No | Yes| No |    | No |    | No | No | 30k|   0|
Stanford   | No | Yes| Yes| Yes|    | No |    | Yes| Yes| 50k| 100|
SUN        | No | Yes| Yes| Yes|    | Yes|    | Yes| Yes|    | 300|
UB         |    |    | Yes|    |    |    |    | Yes| Yes|    |    |
Wollongong | Yes| No |    |    |    |    |    |    |    | 25k| 395|

    The "max FTP" column is for the fastest FTP to a Pc (*not* from)
seen by a user  (in Kb/sec).  It makes  no difference in this  table
which machine was at the other end (obviously the faster the machine
at the other end - the better).

    The  following  table  lists  the  most  popular  Ethernet cards
available and whether the Pc/Ip implementation works with the stated
card.

Vendor      3com  Excelan  Interlan  UB    WD   3Com  UB NIC
            3C501 EXOS205  NI5010   2273A 8003  3C523  PS/2
-----------+-----+--------+--------+-----+-----+-----+------+
Beame      | Yes |  No    |  No    | No  |     |     |      |
CMU        | Yes |  No    |  Yes   | No  |     |     |      |
Excelan    | No  |  Yes   |  No    | No  | No  | No  | No   |
FTP        | Yes |  Yes   |  Yes   | Yes | Yes | No  | No   |
IBM        | Yes |  No    |  No    | Yes | Yes | Yes | Yes  |
KA9Q       | Yes |  No    |  No    | No  | No  | No  | No   |
MIT        | Yes |  No    |  Yes   | No  |     |     |      |
NCSA       | Yes |  No    |  No    | No  |     |     |      |
Stanford   | Yes |  No    |  No    | No  | Yes | Yes |      |
Sun        | Yes |  No    |  Yes   | Yes |     |     |      |
UB         |     |        |        |     |     |     |      |
Wollongong | Yes |        |        |     |     |     |      |

    This table is  called the "Nice  to Have" table.   The functions
listed  here  are  not  mandatory   but  are  useful  in  a   Tcp/Ip
environment:

Vendor      name time fing whoi  NFS gate srce Net  ping SLIP
            srvr                     way  code BIOS
-----------+----+----+----+----+----+----+----+----+----+----+----+----+
Beame      | Yes| Yes| Yes| No |    |    |    |    | Yes| No |    |    |
CMU        | Yes| Yes| Yes| No | No | No | Yes|    | Yes| Yes|    |    |
Excelan    | No | No | No | No | No | No | No | Yes| No | No |    |    |
FTP        | Yes| Yes| Yes| Yes| No | Yes| No | No | Yes| Yes|    |    |
IBM        | Yes| Yes| Yes| Yes| No | Yes|    |    | Yes| No |    |    |
KA9Q       |    |    |    |    |    |    | Yes|    | No | Yes|    |    |
MIT        | No |    | Yes|    |    |    | Yes|    | Yes|    |    |    |
NCSA       | Yes| No | No | No |    |    | Yes|    | Yes| No |    |    |
Stanford   | No | Yes| Yes| Yes| No | No | No |    | Yes| No |    |    |
Sun        | Yes| Yes| Yes| Yes| Yes|    |    |    | Yes| Yes|    |    |
UB         |    |    |    |    |    |    |    |    | Yes|    |    |    |
Wollongong |    |    |    |    |    |    |    |    |    |    |    |    |

FTP Software is OEMed to BICC ISOLAN, Fibronics, Proteon, cisco, Scope,
Spider Systems and Micom-Interlan.

clements@lf-server-2.BBN.COM (Bob Clements) (03/11/88)

In article <8803100323.aa24343@Louie.UDEL.EDU> HANK@BARILVM.BITNET (Hank Nussbacher) writes:

<                         The Pc/Ip Scorecard
<                        revision 3:  03/10/88
 ...
<Vendor      name time fing whoi  NFS gate srce Net  ping SLIP
<            srvr                     way  code BIOS
<-----------+----+----+----+----+----+----+----+----+----+----+----+----+
<KA9Q       |    |    |    |    |    |    | Yes|    | No | Yes|    |    |
                                                      ^^^
						      Yes

This was added a couple months ago. I sent Hank a msg to this effect, but
I guess it didn't reach him.

/Rcc
clements@bbn.com

jerry@oliveb.olivetti.com (Jerry Aguirre) (03/12/88)

Many of the fields such as "ftp", "ping", etc. can have four possible
answers, NO, CLIENT, SERVER, and YES.  For example, many of the packages
support FTP as a client but don't support a server.  This can be an
important distinction.  The alternative is to assume a yes means a
client and have separate fields for the various servers.

ROMKEY@xx.lcs.mit.EDU (John Romkey) (03/13/88)

Here are some comments on the PC/IP scorecard revision 3.0. I think
this is a very useful thing to finally have done, and I would've sent
in some comments before this but I've been pretty busy lately so I
haven't had time till now. When the scorecard is fairly complete you
might want to try getting it in ConneXions, the TCP/IP
Interoperability Newsletter. Talk to Ole Jacobsen
(ole@csli.stanford.edu) about that. 

>    All Pc/Ip implementations support IP, TCP, FTP and Telnet.

No, they don't, and now there's a column for FTP in the chart, so I think
this text needs to be changed slightly.

>Vendor      TFTP  POP ICMP SMTP VT-  3270 FTP   ARP  UDP max  cost
>                                 100                     FTP  ($)
>-----------+----+----+----+----+----+----+----+----+----+----+----+
>MIT        | Yes|    | Yes|    |    |    | No | Yes| Yes|    |  50|
		   NO	     NO   NO   NO    		   5kbytes/sec

>    The "max FTP" column is for the fastest FTP to a Pc (*not* from)
>seen by a user  (in Kb/sec).

Two things should be clarified here. "Kb/sec" is the usual notation
for "kilobits per second" while "KB/sec" is the usual notation for
"kilobytes per second". I read the text as saying kilobits per second,
but I think some of the numbers written there are really kilobytes.
When I'm afraid that I'll be ambiguous, I usually write it as
"KBytes/sec" or "Kbits/sec". 

Also, FTP to what? Floppy disk, hard disk, RAMdisk or the NUL device?
I suggest the NUL device. That removes the disk performance as a variable
affecting the speed. The two other biggest variables are the network
card and the processor speed (for instance, a PC/XT's going to be a lot
slower than a 386 for a several reasons due to the speed difference).
This really is important for the speed column to have any meaning.

To gauge raw TCP throughput, FTP to the NUL device is one good way,
but the best way is to take two of the TCP's on roughly the same
configuration machines and have one of them send data over a TCP
connection to the other as fast as it can and measure the transfer
rate. The receiver just throws the data away when it receives it. This
gets the filesystem out of the way, too, and gives a better idea of
raw TCP throughput. Having the FTP number there does seem to be a good
idea, too, though, since it will be a lot closer to what users will
really see. 

Also, all TCP/IP implementations for ethernet should support ARP. If
one doesn't, it's next to useless since it won't be able to talk
properly to other hosts on the network. And ICMP is another must for
proper internetwork operation - no ICMP and you can't process ICMP
redirects and do proper routing. 

>Vendor      3com  Excelan  Interlan  UB    WD   3Com  UB NIC
>            3C501 EXOS205  NI5010   2273A 8003  3C523  PS/2
>-----------+-----+--------+--------+-----+-----+-----+------+
>MIT        | Yes |  No    |  Yes   | No  |     |     |      |
					    NO	  NO    NO

There are other network interfaces people are probably interested
in, too, especially IBM Token Ring. There are also the Proteon 
p1340 and p1344 IEEE 802.5 (ie: IBM) token ring cards and ProNET-10
p1300 board. You might want to create a category of additional network
interfaces. If you do I'd move SLIP there. Some people might also
be interested in SLIP. Also boards like the Micom-Interlan NI5210 and the
3COM 3C503. Some people like StarLAN too. Micom-Interlan has a StarLAN
version of the NI5210; Western Digital has one of the WD8003. AT&T has
a StarLAN only board.

Another point is that some implementations have very good
interoperability with proprietary LAN systems like 3COM 3+, Novell
NetWare, Banyan Vines or Univation Lifenet. There should be some way
to indicate what software has such interoperability features.

>Vendor      name time fing whoi  NFS gate srce Net  ping SLIP
>            srvr                     way  code BIOS
>-----------+----+----+----+----+----+----+----+----+----+----+----+----+
>MIT        | No |    | Yes|    |    |    | Yes|    | Yes|    |    |    |
	     YES   YES       YES  NO   NO        NO        NO

Questions: what does 'gateway' mean? Does it mean that the software can
function as an IP router? If so, FTP Software's PC/TCP needs a 'no'
there. If it means that the software will work with IP routers, then
MIT's needs a 'yes'. This should be clarified in the discussion.

Also, it should be clear whether 'name srvr' means IEN 116 or Domain or both.
There's a very big difference, and for a TCP/IP implementation to be useful
on the Internet, it almost has to support the domain name protocol. But
it doesn't have to in order to be useful on an isolated LAN.

On source code, does this mean 'free' or 'included' with the binaries, or does
it indicate whether or not the source code is available at all? For instance,
FTP Software's entry says NO. But FTP is willing to sell source, it's just
expensive. There's a big difference between expensive and not-at-all. Maybe
there should be three indications here: "inc" (included), "$" (more money),
"no" (not at all).

A lot of people will probably also be interested in knowing whether a
package has any Berkeley extensions (like rlogin, rexec, rsh, rcp) and
whether it supports servers (FTP, SMTP and TFTP are common).

A useful appendix to this scorechart would be a list of sources of
implementations; how to contact them (US Mail, email, phone, telex,
fax) or get the code over the net (anonymous FTP or Bitnet LISTSERV).
A list of network hardware vendors and contact methods would also be useful.
Between all of us on this list we ought to come up with a complete one.
To seed things off, to get MIT PC/IP, contact:

	MIT Microcomputer Center
	Room 11-209
	77 Massachusetts Avenue
	Cambridge, MA  02139

	telephone: (617) 253-6325

And FTP Software is:
	FTP Software, Inc.
	PO Box 150
	Kendall Square Branch
	Boston, MA  02142

	phone: (617) 868-4878
	telex: 981970
	fax: (617) 864-3943
	email: info@FTP.COM


			- john romkey
		...harvard!spdcc!kaos!romkey
		       romkey@kaos.uucp
		    romkey@xx.lcs.mit.edu
-------