PE> So?  It isn't properly autobauding on ATZ, like I said 
PE> above. Fucked by design.  

DD> It does autobaud.  If you have your modem's NVRAM set to 57,600 and initi 
DD> it at that speed, the change your software to 38,400 and issue another ATZ 
DD> it will autobaud on the AT and send you the OK.  It will then execute the 
DD> ATZ command wich will reset the current speed back to that stored in NVRAM.

It executes the command AFTER sending "OK"?  Very strange.  So
where is that documented, and where is it documented that this
overrides the autobaud?

PW> Although, I am not dead sure that 
PE> it is fucked by design, because the behaviour of the 
PE> autobauding is not documented at all.  If they had said that 
PE> the autobauding doesn't work after an ATZ, then it would 
PE> definitely be fucked by design.  If they didn't say that, 
PE> then it would be a bug. Your call. 

DD> Please quote the part of the manual that speaks of "autobauding"

I will concede this point.  If you wish to say that the USR does
not document that it is capable of autobauding, so I have no right 
to be complaining about ATZ not autobauding then you are correct,
there is no design fault, and no bug.  What it does mean is that
the USR does not have autobaud capability.  Which means it's fucked
for the price you pay for it.  Do you wish to claim that USR does
not autobaud?  BFN.  Paul.
@EOT:

---