SB>>> Can anyone tell me whether it is neccessary to enable 
SB>>> retraining at both ends of a link to actually enable it, or 
SB>>> just one end?

BG>>> As BOTH modems must agree to a retrain, the answer should 
BG>>> be obvious.

 LD> Not really.

 RB> OK, Would YOU answer this for me..why is it then 
 RB> that the 24000 maestro and
 RB> early versions of the 28800 maestro ALL hung up on 
 RB> you when you requested
 RB> a retrain (from any brand of modem other than a maestro)
 RB> Yet BOTH modems were set with retraining enabled.

 
This buggy modem doesn't apply to either side of the argument.
 
Let me explain my reply to Bill fully.
 
I contest that the answer is not obvious because it depends on the modem
at either end of the link.
 
Having retraining disabled on one end of the link does not imply that
this modem will not agree to a retrain if requested as Bill thinks.
 
Bill is correct if he restricts his discussion to the Courier.  His
answer is not universaly correct however,  as I'm using a modem right now
which disagrees with him.
 
How a modem chooses to implement retrain behaviour is up to it.
 
The Courier will refuse the request,  and cause the remote modem to hang
up.
 
The NetComm will accept the request and keep alive.
 
The buggy Maestro will hang up regardless.
 
You, Bill and Dave Drummond all think that retrain disabled implies not
agreeing to requests.
 
It depends entirely on the firmware implementation,  and since the user has
no access to the retrain settings other than via the firmware that is all
that matters.
 
Your modems definition is no more correct than mine.

Ludi.
ludi{at}triode.net.au

--- Maximus 3.01