RS> In fact we KNOW that USR has deliberately chosen to play
RS> clever buggers with the initial handshaking and negotiation
RS> between the modems to try to get a fast connect in V32bis mode.

BG> S34.1=1 should disable the proprietary V.32bis fast-connect in the Courier.
BG> Why not ask Paul to run that for a few days and see if it makes a differenc

RS> I did try that and it didnt help. 15 calls in a row.

RS> It is now clear that for some nutty reason the Couriers alone choose to try
RS> for a V32terbo session, immediately after getting a perfect V32bis connect.

RS> Its less clear what real V32terbo connects do fine detail wise to Couriers.

DH> For some time, Paul's setup did 19200 quite happily into my Courier. 

Sure, I didnt mean that. I meant whether they do it by an initial fast
connect and an IMMEDIATE full retrain very quickly after the CONNECT. Its
quite clear that USR has put a lot of effort into a very fast initial
connect, and its quite possible that that is a byproduct of that approach.
And that its just too stupid to realise that that makes no sense on a V32bis
connect, or the V32terbo protocol means that there is no way the Courier can
know if the other end is V32terbo capable except to demand a retrain
straight after a connect in that quick connect mode, to see if its V32terbo.

There is some hint of that in this,

S34       0          Bit-mapped register.

                     Bit    Value   Result

                     2       4       Disable the faster retrains that occur
                                     during proprietary V.32 terbo modulation.
                                     Used for troubleshooting.

We should really have tried that one too.

DH> So the facility works (although that was 3 SDL revisions ago - it could
DH> well have been improved/broken/removed/upgraded out of sight by now..:-)

I'm not convinced, the other possibility is that V32terbo is a real
kludge and that there are inevitable deficiencys coexisting with the full
range of V32bis modems already in the field before it was even invented.

THAT possibility is supported by the fact that Paul had to
suppress the V32terbo mode on the Thunder because of some problems
with incoming V32bis calls. Not just from a Supra either.
@EOT:

---