David, at 08:15 on Mar 16 1996, you wrote to Paul and Bill...

(Or perhaps that should read Wallace and Gromit...)

DD> After 12 hours of running the M34F the following can be observed:

I'm more interested in what CAN'T be observed, such as AfterCall stats.

DD> ATI5 returns "NetComm Ltd." instead of a list of the data
stored in NVRAM. 
DD> Is this a bug?

Oh, undoubtedly.  There's absolutely no question about it, in fact.

DD> It gives the list on the all of the USRs I've tried . . .

Yep, no argument with that at all.  Ergo, the M34F must have a major bug.

DD> I have a mail downfeed out on the Darling Downs.  He has some sort of 
DD> Rockwell based V.FC modem.  When he rings the Courier or the Sportster he 
DD> usually gets a 19,200 V.FC connect. 

Looks rather bad from the POV of his lines then, I'd suspect.

DD> Files transfer at around 100% (according to Bink 2.59).  This morning he 
DD> connected at 28,800 V.FC, however the transfers only managed 30-35% 
DD> efficiency.

Very interesting indeed, and although I suppose that whilst it's reasonable
to expect a native Rockwell to better implement its proprietary V.FC
protocol with another of its breed, I don't really buy that for one minute.
 Let's basically just cop-out and suggest that his lines were fucked...

DD> Regarding ATZ resetting the port speed. . .

Really Dave, must you resurrect that subject again ?  :)

DD> ATI5 on the USRs lists the NVRAM settings.  These settings include the port 
DD> speed (there it is at the top left of the list).

You mean it DOESN'T on the NetComm ?  No matter, it probably just need
flashing with the latest SDL.  BTW, how does its Distinctive Ring and CID
functions both work?  Are they anywhere near as good as the Courier's
versions ?  :)

DD> AT&V (what sort of a command is that?) will list the NVRAM
settings in the 
DD> M34F 

A standard generic Rockwell command, meaning Very big list of registers...

DD> - there is no sign of any stored port setting to be reset by ATZ . . .

Better ask Paul then, as he understands these things.  :)

DD> Other than the above, I wouldn't think any of my callers would have spotted 
DD> the difference (so far)

Oh really?  Try this little lot for size then...

---------------------------------------------------------
V.34+ Sportster to:      NetComm M34F      USR Courier
---------------------------------------------------------
Modulation               V.34              V.34
Carrier Freq (Hz)        1959/1959         1829/1829
Symbol Rate              3429/3429         3200/3200
Trellis Code             64S-4D/16S-4D     64S-4D/64S-4D
Nonlinear Encoding       ON/ON             ON/ON
Precoding                ON/ON             OFF/OFF
Shaping                  ON/ON             ON/OFF
Preemphasis (-dB)        6/4               4/4
Recv/Xmit Level (-dB)    30/14             25/11
Roundtrip Delay (msec)   3                 3
Blocks resent            0                 0
Retrains Requested       0                 0
Retrains Granted         0                 0
Retrains Granted         0                 0
Line Reversals           0                 0
Blers                    0                 0
Link Timeouts            0                 0
Link Naks                0                 0
Data Compression         V42BIS 2048/32    V42BIS 2048/32
Equalization             Long              Long
Fallback                 Enabled           Enabled
Protocol                 LAPM 244/6        LAPM
Speed                    28800/28800       28800/28800
Last Call                00:00:34          00:00:30
Disconnect Reason        DTR dropped       DISC Received
----------------------------------------------------------

The Courier uses a lower Symbol Rate and Carrier Frequency for the same
link rate, 64-state trellis coding, receives at 5dB better than the M34F,
and transmits a 3dB stronger signal.  A rather interesting old clich‚ comes
immediately to mind here... "Oils ain't oils".  :)

Regards, Bill

--- Msgedsq/2 3.20