BG> 1. What EXACTLY does ATY11 measure ?

JF> It measures the RX levels of a set of specific frequency tones which
JF> are sent during the V.34 handshake sequence. These RX levels are used to
JF> determine the frequency response characteristics of the channel, which
JF> in turn affects V.34's selection of carrier frequency and symbol rate.

BL> Aha! Guessed it!

Poor old Bob. Talk about an absolutely classic example of the senile with
their dicks in their hands making complete fools of themselves. Again.

BL> Bloody USR actually plots the response
BL> while it's shaking hands. Clever buggers.

Poor old Bob. Just goes to show what a complete fool you
can make of yourself bullshitting from utter ignorance.

BG> In fact, the Rockwells do that as well (or are supposed to).

V34 cant work without it. In fact its the whole point of it. Poor old Bob.

BG> That's the whole idea with V.34, and how they manage to wring
BG> the last available bit from a given line.  The real bonus is
BG> that USR actually make this (and other) data available to the
BG> user, while the Rockwells just make you guess.

True, its a very worthwhile feature to be able to see the raw data.

BG> 2. Why is there such a VAST difference between USR & Rockwell ?

JF> That depends on what kind of difference you mean. If the rolloff is the
JF> same general "shape", but the overall levels are
different, that would
JF> indicate that the remote modems are using different transmit levels.

BL> Ahh... two out of two. The pensioner strikes back.

Pensioner ends up face down in the mud, AGAIN, with
muddy footprints all over his back, AGAIN. Poor old Bob.

BG> And explains the reduced Tx levels from your average Rockwell.

You dont even KNOW that they HAVE a reduced TX level Bill. The USR
stats CLAIM they do, and that claim is pretty fucking dubious in fact.
Most obviously the fact that deliberately setting the TX level in
the Rockwell doesnt produce any corresponding change in what the
USR at the other end claims its TX level is. Unless you independently
MEASURE the TX level the Rockwell is using, and determine that when
you have told it to use -11dBm, it actually uses say -17dBm, you dont
actually KNOW that the Rockwells have a reduced Tx level at all.

AND, when we KNOW that one of the Courier Austel SDLs DID have
a significantly reduced TX level, without any dramatic effect on
the performance, the proposition is pretty fucking dubious indeed.

JF> If the rolloff "shape" is different, that would indicate that
JF> individual calls are being routed through different circuits, or
JF> the remote modem is having a v.34 parameter negotiation problem.

BL> I like the sound of this Joe... he agrees with me!

Pathetic really.

BG> Bob and Rod both hate his guts.  Reckon he's a USR zealot.  ROFL!

I have never said he is a USR zealot, just that he gave you the
classic brushoff when you pointed out that there is something
pretty fucked in the USR stats when a Rockwell is called.

Its actually absolutely classic manufacturer's rep finger
pointing. 'cant possibly be us, must be their fault, fuck off'

BL> BTW, I still suspect that USR has a
BL> clever trick they're keeping quiet about.

Poor old Bob, hasnt even grasped what V standards are about. Pathetic.

And USR certainly havent kept quiet about V34+, screamed it to the rooftops.

BL> The 6dB difference sounds a bit much to me for plain transmit level.

Its completely silly, it cant be real. Specially for what
purports to be the STANDARD test for the line probe.

BL> They'd be measuring their "levels" after the signals have been
BL> processed, and any clever processing would show a higher level,
BL> which is what I suspect is happening... plus a few extra volts too.

Silly senile fart, hasnt got a fucking clue what the line probe is
about. Cant even grasp the fact that you HAVE to use a standard test
to actually measure the line to decide what to do with those params.

No wonder the T29 was such a heap of shit.

BG> Dunno, I'm not prepared to speculate,

The fact that V34 does attempt to MEASURE the line characteristics
by putting a STANDARD test on the line and seeing what comes out the
other end is completely beyond dispute, that the whole POINT of V34.

BG> but it's clear that a USR-USR link is superior to any other
BG> combination. Nobody has ever been able to dispute that.

Fraid so. If you eliminate the >28800 speeds, which the M34F doesnt
even CLAIM to support, you havent been able to show that on your
calls to Paul that a USR/USR pair does better than an M34F/M34F pair.

So you havent got a SHRED of evidence that the USR/USR is better there.
@EOT:

---