PE> It depends on x of course, but I would imagine 2 or 3 would 
PE> suffice (considering it has only sent one block, thus we 
PE> don't have to wait for the modem to flush its buffer). My 
PE> calculations say that there should be a saving of 18 seconds 
PE> by going to V42, so the loss of 3 seconds would still be 
PE> substantial savings.

 RS> The problem is that often it comes back and you have to 
 RS> repeat it multiple times in a single session.

I don't believe this happens on V42.  MNP4 yes, but not V42.  Perhaps next
file, but not same session.

 RS> It might get a bit tricky detecting it in Binkley too. The 
 RS> double block never gets out of the modem, so Binkley wont 
 RS> see it. All you have to go on is the time between blocks and 
 RS> that can be higher than usual due to a perfectly valid error 
 RS> recovery between the modems on a real line error or just a 
 RS> pause in the sending system.

No, I was always going to pause after block 0, given the above.  BFN.

Paul

---