PE> 1. The problem may only be seen on the combination of line conditions
PE> seen with both you and me.  I am not sure that we have enough test
PE> results to know whether this problem only happens on 2 bad lines, 1
PE> bad line, or 0 bad lines.

BG> I'm positive that my lines were a contributing factor.

Bill, any credibility that used to be behind one of your "positives"
disappeared a couple of months ago.

I will see if Rod is going to pipe in with any evidence backing a
line factor.

PE> 2. When the problem was happening, it was a sign of AT LEAST 1 modem
PE> violating protocol, and BOTH modems having deficient code.

BG> I'm also positive that BOTH the NetComm AND the earlier Sportster ROMs had 
BG> less than perfect code.  The new ROM fixing the problem virtually proves 
BG> that.

Uh oh, another Bill Grimsley positive.  In this case it is lax
enough to cover what I already said, which is correct anyway.
If you'd said "buggy" instead of "less than perfect" I would have
popped in a snide remark.

PE> 3. USR either stopped violating protocol, or they provided more robust
PE> code.  Most likely the former. 

BG> Most likely both, actually.  The former being the result of the latter.

Of course, by definition.

PE> Only they and Netcomm can answer that for sure, and they are both a bunch 
PE> of wankers so we'll never know.

BG> Perhaps, but I know who makes the better modems.

Bill, the day your "knowing" stopped being worth the paper it was
written on, expired months ago.  BFN.  Paul.
@EOT:

---