Hi folks. I had an opportunity to test a Spirit Viper for
2 days a year ago. I only had two confirmed problems in
that time:
1. I was getting "Too many errors" message in binkley.log.
This turned out to be a binkley bug, not a Spirit bug.
2. I was unable to get the modem to give a 1-line connect
string, despite using the command as documented in the manual.
I did not spend a lot of time on this problem though, and I
could easily have made the mistake my end, the modem was
whisked off me too quick to say for sure.
There were also a couple of people muttering about some
connect speeds, but nothing confirmed by them so should be
ignored. However, the 2-day trial period writes that off
as a definitive test anyway.
Here are some messages from my messagebase...
Date: 1995-01-31 21:16:32
From: Paul Edwards
To: All
Subj: spirit viper
Attr: local scanned
Area: LOCSYSOP
I've got a Spirit Viper online at the moment. Please inform me
of any and all problems so that I can decide whether I should
be buying one or not. BFN. Paul.
Date: 1995-02-02 09:23:42
From: Paul Edwards
To: All
Subj: Spirit Viper
Attr: local scanned
Area: LOCSYSOP
The Spirit Viper is now offline and we're back to the Spirit
Thunder set to a maximum of V32bis. I have my entire Binkley
log for the period available for FREQ. The file is VIPER.ZIP
and the size is 11841 bytes. For those of you with logs or
memories, could you extract the bits that were from you, and
provide some comment as to what went wrong. I will collate
your comments into a file, which I will also include in the
VIPER.ZIP archive. I am using a modified version of Binkley
that prints a lot of debugging info. At one stage I reverted
to the old Binkley which is why the debug comments
disappeared. I don't have a great deal of time, so if you can
try to do any analysis yourself, that would be good. BFN. Paul.
Date: 1995-02-05 14:08:18
From: Paul Edwards
To: All
Subj: too many errors
Attr: local scanned
Area: LOCSYSOP
I know the reason for the "too many errors" error message I
was getting from Binkley whilst I had the Viper. There is
a bit of code that sets a variable n, which is an integer,
to the number of characters that may come in before getting
the header. n is set to the baud rate, which is an unsigned
integer. From the Viper, I was getting a connect string of
38400 (unfortunately, despite me trying to change that).
When this was converted to an integer, it became negative.
The program was subtracting 1 from n, looking for it to go
negative, then display the error. n is now a long!
BFN. Paul.
@EOT:
---
* Origin: X (3:711/934.9)
|