🚀 go-pugleaf

In article <WOHL0BAS1lhyEwKy@ilf0rd.demon.co.uk>, Keith Stanbury
<KEITH@ilf0rd.demon.co.uk> writes (snipped)
>The following, gleaned from alternative ISP's which also uses Ascend
>The main connection problems appear to centre around the V42
>compatibility between the ASCENDS and receiver modem. This requires
>work
>on the ASCEND end - if problems on connection persist, one of the
>following *may* prove a workaround -
>
>ii)             AT&F%C0\N\V4   (disables V42 compression/error
>HTH
I have been with DEMON for 12 Months and was using the original windows WINDIS
(DISWIN) distribution using the latest version of Trumpet Winsock until my
recent purchase of TurnPike. I now use the NTS winsock provided with TurnPike
3.01. I have had a persistent problem with my Motorola 3400 showing up to 2000
overflow errors per hour under Trumpet Winsock and about the same with the NTS
winsock. I use the software for both work and pleasure and the machine specs are
:-
 Work  -  Compaq 486 DX2/66 16Mb Ram Seagate
          420 Mb Disk 16550 uart serial
 Home  -  Dabs 486 DX2/66 24Mb Ram Seagate 540&1200 Mb Disks 16550 uart serial
          using the most recent seagate 32bit disk driver from their support
          page. 
My machine at home is (by feel) somewhat faster than the one at work (probably
because it is VLB using the fast seagate 32bit file disk driver).   
The main difference is that I use a US Robotics 28.8 Sportster at work and a
Motorola 3400 28.8 Online at home. As far as I know I have never had a single
overflow error on the machine at work. This leads me to support the faction that
believes there is a compatibility issue between the Motorola and the Demon
Ascend setup. I can only get below 10 errors/Hour if I disable V42 on the
machine at home. (%C0\N as per Keith Stanbury's suggestion) (I didn't find &G2
made any difference?) This gives me a typical max speed of circa 5.5
Kilobytes/sec at work and 3 kilobytes/sec at home. Doesn't say much for the
Motorola does it? (unless anyone can think of something I might have missed (yes
I do use COMxFIFO=on on both machines with a suitable buffer and I don't think I
have a cable fault)

-- 
David Moor