;
In a msg of , Roy J. Tellason writes to Peter
Knapper:
;
Roy,

 PK>> I am curious. Can you please explain how a communications
 PK>> driver that handles the V.24 H/W interface for the Operating
 PK>> System, can possibly affect the Analog interface of a Modem?
 PK>> Its news to me that the serial port is in any way involved
 PK>> in the Analogue line negotiations that the modem makes with
 PK>> its Analog connection point...

 RT> It would seem to me that a port driver that's better written and
 RT> able to deal with actually moving more data through a port would
 RT> also support higher connect rates...

Roy,

With all due respect, that's disappointing... coming from you.  While the
first part of what you say is quite correct, there is simply no way that a
port driver, regardless of how well written, will affect the *analog
negotiation* of connect speed.  At the time the analog
negotiation/handshake/whatever you'd call it takes place, the modems don't
know what they are connected to, or what driver is controlling the ports, nor
do they care!  The negotiation of connect speed, and any subsequent shifts up
or down as required, is entirely conducted between the modems, without respect 
to the quality of the drivers or the serial port hardware.

Now, it is possible for *flow control* (Xon/Xoff or RTS/CTS or both) to affect 
*throughput* due to differing driver quality, but that is unlikely on anything 
newer than a 286 machine (even with 16450 UARTS) and that has nothing
whatsoever to do with the negotiation of *connect speed*.

For an interesting test, it is possible to get two external modems to
negotiate a connect when only one of them is connected to a computer! At least 
it will work with externat USR modems... I've done it. Obviously, the callee
must be the one that is not connected, but the caller WILL dutifully report
the connect speed.

-[Steve]-

--- GoldED/2 3.0.1/#