Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!clyde.concordia.ca!uunet!bywater!acheron!archet!larouch!jparnas
From: jparnas@larouch.uucp (Jacob Parnas)
Newsgroups: comp.dcom.modems
Subject: Re: Microcom Modems.....Are they any good?
Message-ID: <1989Dec15.082753.709@larouch.uucp>
Date: 15 Dec 89 08:27:53 GMT
Reply-To: jparnas@larouch.UUCP (Jacob Parnas)
Organization: Larouch Headquarters, Hopewell Junction, NY
Lines: 45


In article <7134@ki4pv.uucp> tanner@ki4pv.uucp (Dr. T. Andrews) writes:
>In article <1989Dec4.054323.9528@larouch.uucp>, jparnas@larouch.UUCP (Jacob Parnas) writes:
>) ... we use the Microcom's ... have the best compression currently
>) around ... hook them up ... at 38400 baud.  I've seen up to 33000
>) baud true throughput cating a file like /etc/hosts. You won't get
>) that from a T2500.
>Quite true.  From this, I believe that we can judge what the best modem
>is when we wish to call up and cat /etc/hosts.

Obviously, buying a modem only for catting /etc/hosts is stupid.  The
reason that I noted the throughput of catting /etc/hosts was that I wanted
to test the throughput of the modem in an interactive ascii environment.
/etc/hosts was a large file (~140Kbytes) that I had on hand.

I (and I believe many other people) like to do interactive ascii computing
over a modem (from home).  This includes applications that require sending
data that is easily precompressed.  For instance, editing files or reading
news requires that data be sent without precompression, and it is very
nice for it to be extremely fast.  I find that if I have a large window
(say 60 lines), it is much nicer if a screenful of a usenet article is
sent to my screen at 33000 baud instead of 19200 baud. 

There are many other applications that it is nice to have compression built
into the modem.  For instance, SLIP.  I notice a big difference between xmh
performance with the Microcom's MNP 9 when compared to the Telebit T2500.
Some other applications which benefit from MNP level 9 are: ftp, rcp, telnet,
xrn, rdist, and 3270 emulation to VM/CMS systems.

>On the other hand, if we are pushing around large amounts of compressed
>data (a newsfeed is a convenient example) where the modem's compression
>is not effective, then you should consider a modem whose underlying
>transport technology is better suited to moving bits.

I agree that MNP level 9 is not useful for sending precompressed data.  I use
a Telebit to get usenet and mail.  But, for interactive communication of
data that isn't precompressed, MNP level 9 is extremely nice.  

------------------------------------------------------------------------------
| Jacob M. Parnas                    | DISCLAIMER: The above message is from |
| IBM Thomas J. Watson Research Ctr. | me and is not from my employer.  IBM  |
| Arpanet: jparnas@ibm.com           | might completely disagree with me.    |
| Bitnet: jparnas@yktvmx.bitnet      \---------------------------------------|
| Home: ..!uunet!bywater!acheron!larouch!jparnas | Phone: (914) 945-1635     |
------------------------------------------------------------------------------