-=> Quoting Bill Cooper to Rick Collins <=-
-=> FidoMail to 1:163/215, please.-=<
bc> Ok... so what's your point, or, what's the big deal? The reality
bc> of this matter is that no one really cares what you, or, I
bc> think... nor will it matter in a few months. 
I'm sure others who wonder why they can't achieve irrational
throughputs appreciate the information that explains how such
throughputs are possible - and what they mean.
-> If you want to believe it wa because it was early in the morning,
-> fine.  If you want to believe that established fact is nonsense,
-> fine.
bc> In this area, the lines seem to be "cleaner" early in the
bc> morning, thereby allowing for a faster connect speed... when
bc> traffic is heavy... It would appear that line quality suffers.
bc> Your interpretation of my statement is insufficient for you to
bc> quantify your obscure answer. 
Your max connect speed is 33.6 with V.34.  With that maximum speed,
and no modem-modem compression (your original claim) you _cannot_
exceed about 4070 cps.  You claimed over 4400.  That means one of the
assumptions is wrong.  Here are the assumptions:
1) Max connect speed with V.34 is 33,600 bps.
2) Max modem-modem throughput at 33,600, with error correction is
4200 bps.  EC overhead reduces this to the 4000 cps range.
3) The files were fully compressed and could not be compressed
further with V.42bis or MNP 5.
Assumption 1 is incontrovertible.  The modem can't connect faster
that it's designed maximum speed.  Assumption 1 is therefore valid.
Assumption 2 can be easily proved.  With error correction, the modem
transfers characters as 8 bits.  Therefore, in one second a maximum
of 33,600/8=4200 "characters" can be transferred between the modems. 
Some of those characters are involved in the EC overhead, and,
therefore, _less than_ 4200 characters will be passed from the modem
to the PC.  Assumption 2 is therefore valid.
Assumption 3 remains as the only questionable assumption.  Given (1)
and (2), and the fact that you observed "4400 cps" throughput,
assumption 3 cannot be correct.  Ergo, the files were not fully
compressed, and some modem-based compression was in effect.
BC> My children still believe in Santa Claus... does that qualify?
I do to.  But the Great Pumpkin is just too much to accept. :-)
TTFN. Rick.
Ottawa, ON 2 Apr 12:15 
--- Blue Wave/DOS v2.20
---------------