Hello Ludi!

Replying to a message of Ludi Desouza to David Drummond:

 DD>>> If enough modems make the same error, does that make the
 DD>>> error a standard? Remember the Netcomm modems of a few years
 DD>>> back?  Had no problems with other Netcomms....

 HM>> Yes, but there are more Rockwell modems than NetComms -- and
 HM>> more than USRs even.

 DD>> Still doesn't make it the documented standard.

 LD> Umm..  your not going to give in on this one are you?

 LD> There was a message from a Supra tech to a USR tech about a year ago
 LD> which described the problem.  Their opinion was that USR must have
 LD> had a different interpretation of the V.34 specs by choosing 3200
 LD> instead of the 3429 symbol rate when presented with the particular
 LD> set of values in the INFO sequences.

 LD> There was a rumor going around that USR was protecting against
 LD> the early Rockwell V.34 chip revision which had problems at the 3429
 LD> symbol rate.

 LD> Supra and others changed their firmware to work with USR modems,
 LD> USR then changed their firmware to work with the rest of the
 LD> Rockwells.

 LD> I'm not sure if Rockwell themselves have made any changes to their
 LD> stock code.

 LD> Everyone can now interoperate except unmodified Rockwells calling
 LD> Austel Couriers, as they still use the old USR firmware.

 LD> I would have to agree with Hamish on this, if it is a problem with the
 LD> interpretation of the standard then given the state of the modem
 LD> world it should be up to USR to fix it.

I agree that USR have modified their code to make it more compatible with
the Rockwell /bug/ - in the USA.  For some reason they don't seem so
enthusiastic about updating the Austelised code. Rumour has it that it may
be some protectionist racket pushed by the Rockwell based Australian modem
industry.

I have tried the later USA SDLs in my modem.  They definitely *DO* connect
better with Rockwell based modems.

David!

--- FleetStreet 1.19 NR