Path: utzoo!censor!geac!torsqnt!hybrid!scifi!bywater!uunet!hayes!tnixon
From: tnixon@hayes.uucp
Newsgroups: comp.dcom.modems
Subject: Re: Help: Why MNP5?
Message-ID: <3764.27aaef05@hayes.uucp>
Date: 2 Feb 91 16:55:32 GMT
References: <1991Jan18.194613.13435@watserv1.waterloo.edu> <0cD8V3w163w@ozonebbs.UUCP> <1991Jan27.160653.7104@nstar.rn.com> <3455@d75.UUCP>
Organization: Hayes Microcomputer Products, Norcross, GA
Lines: 27

In article <3455@d75.UUCP>, woan@nowhere (Ronald S Woan) writes:

>>Sorry - but incorrect with MNP5.  MNP5 over a 2400 baud connection
>>if properly installed will produce transfer rates of 280 cps even
>>sending GIFs and ZIPs
> 
> According to the periodic FAQ, MNP 5 slows things down compared to
> just running MNP 4 for precompressed files... V.42bis automatically
> detects precompression and lets it go through without the compression
> overhead.

MNP4 is capable of throughput of about 292cps (2400/8 bits per 
character = 300cps, then you subtract for protocol overhead [7 bytes 
of overhead for every 256 bytes of user data]).  MNP5's _average_ 
token length is 9 bits, which means that on _completely_ random 
data, MNP5 sends an average of 9 bits per character instead of 8 -- 
but this is still less than the 10 bits per character of "pure" 
async.  MNP5 would therefore, worst case, send only 260cps on random 
files.  It is rare to find a precompressed file that is "totally" 
random, so slightly better throughput might be possible, but I doubt 
you'd see 280 -- but maybe so.  

-- 
Toby Nixon, Principal Engineer    | Voice   +1-404-449-8791  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       hayes!tnixon@uunet.uu.net