On 08-25-96 Bill Grimsley wrote to Michael Raiteri... 
 
 BG> At the risk of offending some Maestro owners (so what else is new?), 
 BG> my 
 BG> experiences with Banksia's V.34 modems have generally been 
 BG> good, but some Maestros appear to suffer from the most 
 BG> peculiar problems, such as their oft-faulty piezo 
 BG> "speakers" which, for some inexplicable reason, stuffs up 
 BG> high-speed connects.  Their QA could do with some 
 BG> improvement, IMO. 
 
Hmmm thats bad.  My ISP is about to ditch Banksia and Microcom modems 
in favour of a total Maestro setup.  I hope it doesn't affect me too 
much. 
 
 BG> MR> BBS has.  The BBS is 
 BG> MR> QUANTUM LEAP ph 07 3805-1180. 
 BG> 
 BG> There is no mention of the brand in his log-on screens, 
 BG> although the Courier's aftercall stats indicate that it's 
 BG> probably a Rockwell-based V.34 modem, but neither a Spirit 
 BG> Viper nor a NetComm (no LAPM/SREJ).  I'd guess a Maestro? 
 
Believe it or not the sysop finally found time to reply and he said 
that its a Netcom V34 Roadster.  Interestingly enough he said nearly all 
the high rate connects vary between 19200 and 26400. 


 BG> MR> I'm seeing /ARQ connects alright but monly for 25 seconds :( 
 BG> 
 BG> In that case, it could well be that he has retrains 
 BG> disabled at his end (this was unfortunately the default 
 BG> with earlier Maestros, for example), and when your Dynalink 
 BG> requests a retrain, the remote can't respond, then drops 
 BG> carrier.  Adding %E2 or %E3 at his end should fix that 
 BG> problem (if that's the cause). 
 
Does the Netcom V34 Roadster have the same defaults. 
 
 BG> MR> Dynalink couldn't shed any light on it.  Their support reckon they 
 BG> have never had the need to disable the V42 detect phase. 
 BG> 
 BG> They like seeing 28800/NONE connects, do they?  What a 
 BG> bunch of morons.  The simple fact is that some Rockwells DO 
 
That I can't disagree with :) 
 
 BG> have a dodgy implementation of V.42, which is why disabling 
 BG> the USR's V.42 _detect_ phase solves that problem (thereby 
 BG> forcing a non-negotiated LAPM connect).  It's a similar 
 BG> problem to that of some Hayes models with their faulty 3429 
 BG> symbol rate, requiring said symbol rate to be disabled when 
 BG> calling these modems, otherwise they simply won't connect. 
 
BTW what specific Rockwell type modems do need those disabled so I 
can be on he lookout for them. 
 
 BG> To avoid further long-winded comparisons between the 
 BG> Dynalink's and the Courier's S-registers, here are the lot. 
 BG>  Please save this chart for future reference, and refer to 
 BG> S27 in particular... 
 
  <<<<<<<<<<<<<  CHOMP CHOMP
>>>>>>>>>>>>>>>>> 
 
Exactly my point - there are two specific different S registers 
one for Disabling V42 Detect Phase and one for Disable V42. 
Dynalink only has S15=128 =disable V42. 
 
I'm not saying that it's not meant to be disable V42 Detect Phase 
it's just thats its pretty ambiguous. 
 
 BG> I'm buggered if I know either, but there's a pretty easy way to find 
 BG> out; 
 BG> simply call Dave's Courier with S15=128 in your dialling 
 BG> string, and if it connects with LAPM, that bit only 
 BG> disables the detect phase.  If you see an MNP4 connect, it 
 BG> has disabled LAPM only, and if you see no error correction 
 BG> at all, then it will have quite obviously disabled V.42 
 BG> altogether. 
 
I did like you said. I dialled David's BBS with a dial string of 
AT&F1S15=128DT and i got the following connect 
 
  31200/ARQ/V32/MNP/MNP5 
 
How do you interpret that? 
 
Cheers 
Mike 
 
 
 
 
___ 
 * OFFLINE 1.56 

--- Squish/386 v1.11