Subj : binkley xe
To : Neil Walker
From : Archie Swan
Date : Tue Jan 21 2003 02:49 pm
Hi Neil,
You wrote on 20.01.03 about "binkley xe":
[hack to my "playing" bit]
AS>> One probable difference on my call to Roy though, I've got my
AS>> RxWindow in Hydra set to 8K at the moment. Both Roy's and
AS>> your's are possibly using the default full-stream in the
AS>> RxWindow and TxWindow settings ................ until an
AS>> interfering oddball point sneaks in and remotely adjusts same
AS>> behind yer back :)
NW> Hydra doesn't seem to be at it's best in BinkleyTerm (phew, back
NW> on topic). ZedZap seems to work better for uni-directional
NW> transfers.
During the past month or so, I've been keeping tabs on your
Bink's Hydra performance. Definitely "iffy" at times - but it's
not a consistant "iffyness".
Best way to sorta describe it, is probably in terms that Martin
would understand. By that I mean, it's closely similar to the
way T-Mail used to perform until we got the message across to
Andy Elkins and he fixed it.
T-Mail, somewhat similar to Bink's implementation, also had the
Janus core being accessed - and I've a feeling it's something to
do with that.
For those who're interested in the spasmodic "basic fault" -
Hydra timeouts occur at the session end, which causes an error
state and a re-dial action subsequently takes place.
Is there any means whereby you can totally disable Janus with
your BinkXE ?? If so - would you be willing to do so for maybe a
week or so ?
I'll then go back to bashing with Arjen's "standard" (currently
I've disabled that and am using ZedZap, as you *might* have
noticed).
