🚀 go-pugleaf

In article <847555766.14786.0@aster.demon.co.uk>,
spin <spin@aster.demon.co.uk> wrote:
> In article <847449808.17201.0@mbrown.datastream.com>, mikeb@helion.demon.co.uk 
> says...
> >
> >Stephen wrote:
> >> 
> >> I have a prob with my 3400 regarding Netscape 3. If I am downloading
> >> something from the web, I have to run another prog, such as Agent in
> >> the background and be downloading stuff from the newsgroups otherwise
> >> Netscape just stops downloading and the modem hangs. Any ideas??
> >> The init string I am using is AT&F&C1&D2\Q3\V3
> >> 
> >> Stephen
> >> stephen@stephron.demon.co.uk
> >
> >Since you have Netscape running at all I assume you aren't having the
> >ubiquitous Ascend connect/line drop problem.
> 
> 
> I have a Motorola 3400 and am using Win95 dial up. I also use the
> same init string (with the exception of \V2 instead of \V3 - a trivial
> difference). I always connect to ascends.
> 
> I _almost_ never get line drops, and I _almost_ always get a connection
> first time. What I do get though are line freezes. All IP traffic stops,
> but the connection stays up. The only solution is to disconnect and
> redial. With the current News failure at Demon, I have been doing
> more surfing than normal (Netscape 3) and have found that these freezes
> are much more common when WWW surfing than when downloading news articles.
> 
> I had never experienced this problem with Demon's old USRs.
> 
> My guess is that the ppp software on the ascend's is a bit flakey
> (or the motorola/ascend mix is flakey).
> 
> Any advice or comments appreciated.

Is ther any way to get statistics after a call. In particular counts
of the number of errors which needed correction and the number of
times the modems needed to retrain might give an insight. What you are
describing sounds very much like the stalls caused while modems
retrain. It is also possible that the Ascend is giving up during the
retrain and the call is being terminated at the Demon end, even though
your modem has not reported loss of carrier.