RS> I've also had atleast two calls which managed to start
RS> off at 7200, and fell forward to 9600 and then to 14400
RS> thru the call. One on 31-Mar, another at 07:16 19-Mar.

RS> I basically let these go to see what happens, and the M34F is rather
RS> slow to decide it can step up. It does so fine when it decides to tho.

PE> An EQM of 0?!  I dunno, whaddayareckon?

Trouble with EQMs tho is that they are AT the particular speed in use, they
ARENT a measure of the line itself but essentially a measure of the error
rate at the particular speed, measuring the error in the real constellation
bits compared with where they should be in the constellation.

PE> I was thinking that this might be a substitute for the old occurrence
PE> of a failed connect to the Spirit, but instead, the Netcomm succeeds
PE> in stepping down to a more reasonable speed rather than failing.

I doubt it. It looks more like a massive brain fart
given I never did get anything remotely like that with
any other system I ever called at a high freq for mail.

PE> Or would you expect V32bis modems to always connect at 14400?

Certainly not at 7200, thats extremely low. I would expect that if the line
at your end or mine was marginal, we would occasionally see a 12000 connect
and in fact I did with one particular line with the City BBS which I called
daily for mail for a long time. He is on one of the most decrepit exchanges
in Sydney apparently and doesnt have his lines in a rotary group, and you
could see one line being a bit down performance wise. It always connected
fine, sometimes at 12000 and normally did manage to fall forward up to 14400
thru the session. Occasionally it connected at 14400 and fell back to 12000
for a bit and back up to 14400 too.


And this one is unlikely to be due to the voice call dialing coz
once the modems have fallen forward up to 14400, the thruput is fine.
@EOT:

---