Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!samsung!uunet!hayes!tnixon
From: tnixon@hayes.uucp
Newsgroups: comp.dcom.modems
Subject: Re: V.42 questions on T2500 -- How?
Message-ID: <4038.28688adb@hayes.uucp>
Date: 26 Jun 91 12:38:50 GMT
References: <1991Jun17.225901.10900@coplex.uucp> <1991Jun18.123604.9315@hou.amoco.com> <78@uis-oc.UUCP>
Organization: Hayes Microcomputer Products, Norcross, GA
Lines: 45

In article <78@uis-oc.UUCP>, bob@uis-oc.UUCP (Robert J. Mathias Jr.)
writes: 

> Some of us like to squeeze the last ounce of speed out of our equiptment, so
> we poor miss guided souls prefer to disable V.42 and go with MNP4.  

A good V.42 LAPM implementation can go faster than MNP4.  Why?  
Because the maximum frame size of MNP4 is 256 octets.  You have 7 
octets of overhead per frame, and, considering the average 
occurrence of a zero bit insertion every 63rd bit, the maximum 
throughput of an MNP4 modem is 119.7% of a non-error-correcting 
modem (e.g., 1150cps on a 9600bps modem).

But LAPM can negotiate much larger frame sizes, limited only by the 
amount of buffer space in the modem and the effectiveness of the FCS 
chosen (for example, the 16-bit FCS isn't as effective if you send
more than 4096 bytes in a frame).  Hayes modems, for example, 
support up to 512 bytes in a frame, which, if enabled, would give 
you throughput of (1.25*512/519*62/63=) 121.4% of a 
non-error-correcting modem (e.g., 1165cps on a 9600bps modem).

But, as you probably know, the default frame size for LAPM is 128 
bytes (116.6% throughput, or 1120cps on a 9600bps modem).  Unless 
the modem you happen to be using allows you to specify a larger 
frame size, that's what you'll get.  The smaller frame size does 
allow for faster recovery from line errors, should they occur.

> As to
> V.42bis and MNP5, V.42bis is the winner but since I do alot of file xfers
> of ZIPed files, I also disable compression on my modem.

V.42bis allows the transmitting modem to switch between sending 
compressed and uncompressed data.  Most V.42bis modems incorporate a 
performance feedback mechanism, so that the modem detects when data 
is uncompressible and switch to sending the raw data automatically.  
Thus, leaving V.42bis enabled all the time will normally not impact 
on your throughput at all when sending pre-compressed files, but 
will quickly kick in and increase your throughput on uncompressed 
text during interactive periods between files.

-- 
Toby Nixon, Principal Engineer    | Voice   +1-404-840-9200  Telex 151243420
Hayes Microcomputer Products Inc. | Fax     +1-404-447-0178  CIS   70271,404
P.O. Box 105203                   | UUCP uunet!hayes!tnixon  AT&T    !tnixon
Atlanta, Georgia  30348  USA      | Internet       tnixon%hayes@uunet.uu.net