Frank Sexton wrote in a message to Bob Juge:
 DH> Have a look at an interesting article on the x2 /
 DH> K56Flex / v.90 state of affairs at Boardwatch
 DH> magazine's Web page:
 DH> http://www.boardwatch.com/mag/98/mar/bwm24.html
 FS> I've seen it. It says that x2 is much better than
 FS> K56Flex. Doesn't mean much now that v.90 is here.
 BJ> Read it again.
 FS> What are you getting at?
Perhaps that there are implementation details?  Given the noted similarity 
between x2 and K56Flex, you would expect similar performance.  Differences 
could easily be due to one manufacturer doing a better job of implementing 
their hardware and software.  The ITU-T recommendation does not go into the 
details of hardware and software, it is more concerned with the desired end 
result.  Much like the "black box" theory of software design where the 
internal details of a module are hidden with only the expected input and 
outputs being specified.  In the case of V.90, the client modem is 
responsible for quite a few decisions about how the data will be sent, 
decisions that can have a rather large effect on the effective transmission 
speed.  Heck, Rockwell never did get around to implementing 64S/4D trellis 
coding despite it's advantages.
Given that USR/3Com seems to have done better at implementing x2 than 
Rockwell did in implementing K56Flex, it would appear somewhat logical to 
expect USR/3Com to have the edge when it comes to implementing V.90.  A 
difference that is likely to disappear as Rockwell tunes their code -- a much 
easier task with flash ROM.  You might remember (or want to forget) the early 
v.32bis Rockwell chipset modems?  I still have my screwdriver collection -- 
Supra used to ship a combination Phillips/slot screwdriver with the ROMs.  A 
simple phone call to obtain the code and a few minutes to re-flash the modem 
is a much more attractive option.
Regards,
       David
--- timEd/2 1.10+
---------------