|
|
| Author |
Message |
| 25 new of 227 responses total. |
jor
|
|
response 100 of 227:
| 
Jan 17 14:11 UTC 2000 |
Switching to ANSI I get the same message.
After the message elm shuts right down.
|
davel
|
|
response 101 of 227:
|
Jan 17 14:16 UTC 2000 |
Well, your .login contains this line:
setenv TERM ansi
So whatever your emulation is set to, Grex thinks it's "ansi". Try doing
setenv TERM vt100
and then trying elm again. (It was fairly recent that elm was updated to a
new version; I'd guess that this change accounts for the timing of your
problem.)
|
jor
|
|
response 102 of 227:
|
Jan 17 20:37 UTC 2000 |
That was dramatically successful.
Thanks.
|
scott
|
|
response 103 of 227:
|
Jan 18 02:07 UTC 2000 |
For those among us who still use (gasp!) modems, some of the modems
accidentally got turned off until Monday morning (when I dropped by the
Pumpkin and turned them back on).
|
janc
|
|
response 104 of 227:
|
Jan 18 19:12 UTC 2000 |
My mom was often having disconnects vaguely similar to what Keesan
reports. It went like this: she'd dial in, go into the menus, go into
mail, select the "mail" program (she's not a pine user), get the list of
messages, try to read the first one, and be disconnected while in the
headers. She says it happened pretty repeatably. Not all the time, but
often enough to cause her to give up using Grex for some time. After
the disconnect she always got the NO CARRIER message, so the disconnect
was on the modem connection, not between the computer and the modem.
The modem cable is wired for full handshaking. There was no visible
noise before the disconnect.
Recently, I replaced her US Robotics 2400 baud modem with a Practical
Peripherals 9600. At the same time I noticed that the modem cable
wasn't plugged very tightly into the computer, and screwed that down.
The tests I did afterwards seemed OK (I could cat large text files
without dataloss or disconnects), but what with Grex's downtime, my mom
hasn't been using it long enough to be sure the problem is really gone.
|
steve
|
|
response 105 of 227:
|
Jan 18 21:00 UTC 2000 |
I can't see how the tightening of a cable could affect this, if
her problem was in pine when trying to read mail. If it were a
flakey connetion with too high a resistence I'd expect to see garbage
come out at random.
Forgive me for nor going back and reading past responses on this,
but was/is keesans problem at this point as well, in pine trying to
read mail? If it was/is I have an idea, namely to run typescript
and record *every* character coming out of Grex--I'm wondering if
somehow something is getting sent out as a hangup command to the
remote modem...
|
keesan
|
|
response 106 of 227:
|
Jan 18 23:04 UTC 2000 |
I will tell Jim to run typescript, if he knows what this means.
Is it a UNIX program? One of our modems is internal and used to work fine.
The problem was in pine in general - it would disconnect while answering mail,
while reading mail, or simply when I typed pine. And also happened while
typing responses in bbs. Jim thinks pico is involved.
|
steve
|
|
response 107 of 227:
|
Jan 19 00:58 UTC 2000 |
Oops! The program name is "script"--it creates the file typescript
in your home directory.
What it does is to record every character that goes in or out from
you. It can then be used to analyze whats going on during the moment
of disconnection.
If you can tell tell Jim to run that, then run script, that might
provide us with some answers for you. If you run pine and it works,
you want to remember to stop script from running. To do that, at a
shell prompt (after you exit pine) type exit, and it will finish
recording. A classic problem with script is that someone turns it
on, but then forgets about it, building up a gigantic typescript
file, which they didn't want.
|
gull
|
|
response 108 of 227:
|
Jan 19 03:14 UTC 2000 |
A loose cable actually *could* cause a modem to drop a connection with a NO
CARRIER message. Most modems can be configured to disconnect when the DTR
line is dropped; if that connection in the cable was loose, the modem would
disconnect every time it lost contact. For that matter, if you somehow had
things set up so DTR was a handshaking line, it'd disconnect every time you
got a long stream of data faster than the computer could accept it.
|
mcnally
|
|
response 109 of 227:
|
Jan 19 03:20 UTC 2000 |
Certainly a loose cable could cause sporadic "no carrier" errors, but
they wouldn't be limited to Pine sessions..
|
n8nxf
|
|
response 110 of 227:
|
Jan 19 12:26 UTC 2000 |
The disconnects were not sporadic.
|
n8nxf
|
|
response 111 of 227:
|
Jan 19 13:13 UTC 2000 |
My modem connection is acting odd. When I am at the commend prompt, under
bbs, I am able to backspace just fine. However, when I respond to an
item, the backspace key generates a leftarrow[D. I have it set up to
send a ^H. It sends the same string if I do a manual Ctrl H. (At least I
have a flakey internet connection to fall back on ;-)
|
pfv
|
|
response 112 of 227:
|
Jan 19 17:20 UTC 2000 |
s'ok.. Elm suddenly told me it was unable to find my saved messages.. I
had to teach it to look in a totally new place... Or something.
Was a backup written to the drives?
|
steve
|
|
response 113 of 227:
|
Jan 19 22:56 UTC 2000 |
I'm not sure I understand your question there. The contents of
/var/spool/mail/*/* that is on Grex is from a backup, but it *should*
be identical to what was on the previous drive! Perhaps I'm not
getting at what you are saying?
|
pfv
|
|
response 114 of 227:
|
Jan 20 01:42 UTC 2000 |
I use procmail, so I had a short script that gets me to what is saved..
Yesterday it decided to tell me I had no file such & so..
So, I looked around, found the right path and changed the script.. BUT,
the script had been right for months & months, ergo: was a backup
written back on our drive, or was there something else done?
paths or scripts don't, typically, mutate spontaneously..
|
goose
|
|
response 115 of 227:
|
Jan 20 02:03 UTC 2000 |
I'm having the same TERM problem jor mentioned a few responses ago.
I can't find a similar line in my .profile (I use bash) and the only
thing I can think of is that I used the menu system to create
a .yeswrite earlier today....I used elm earlier today, used the menu, and now
elm is hosed.
|
steve
|
|
response 116 of 227:
|
Jan 20 12:57 UTC 2000 |
Pete, as far as I know, nothing was changed. Can you show me the script
such that we can see the path that changed?
|
goose
|
|
response 117 of 227:
|
Jan 20 17:23 UTC 2000 |
something had to have changed...pete, myself, and jiffer have all had related
problems with TERM.
|
prp
|
|
response 118 of 227:
|
Jan 20 22:12 UTC 2000 |
Several months ago now, I tried using 761-3000 with a 2400 baud modem
(Hayes Smartmodem external). I ran into all sorts of problems, including
frequent line drops. My conclusion was that the Grex modems are not
doing the protocol negotiation quite right.
|
pfv
|
|
response 119 of 227:
|
Jan 21 04:37 UTC 2000 |
re 116:
I was using "~pfv/bin/elm" - now "~pfv/bin/myelm"..
All I had to change was the final directory-element, for some
really, REALLY weird reason.
|
mcnally
|
|
response 120 of 227:
|
Jan 21 15:02 UTC 2000 |
When you say "change the final directory element" do you mean the
change from "elm" to "myelm"?
That would suggest a path order problem..
|
pfv
|
|
response 121 of 227:
|
Jan 21 15:37 UTC 2000 |
Is now:
/usr/local/bin/elm -f ~pfv/mail/HOLD $1
Was - and worked right - before:
/usr/local/bin/elm -f ~pfv/Mail/HOLD $1
I have no idea why the change was required, but it works.
Further, I've spent like an hour with goose doodling with his
.login/.profile: suddenly, grex has decided to set his TERM
to ansi - and he was unable to run elm. I told him to just
"hardwire" it for vt100/102 and stop trusting grex to "determine"
it for him.
BUT, the point remains: these are ALL RECENT "ISSUES" that had NOT
been there BEFORE.
|
mdw
|
|
response 122 of 227:
|
Jan 21 18:04 UTC 2000 |
elm was last changed 29 dec 1999; this was probably the y2k upgrade.
It's possible the new version is either linked with a different curses
library, or is somehow doing screen I/O differently; if so, it's
probably a local build configuration not something inherent in elm.
User pfv has "." before "~pfv/bin" in his path; if he were in a
directory that contained a copy of elm (such as /usr/local/bin) that
would be found and executed first.
|
pfv
|
|
response 123 of 227:
|
Jan 22 02:20 UTC 2000 |
Yes, he would.. Except he did not, nor does he.
|
gelinas
|
|
response 124 of 227:
|
Jan 23 02:54 UTC 2000 |
The only difference is the name of a subdirectory in ~pfv: mail
and Mail. He had been using a file in Mail, now he has to use one
in mail. Interestingly, Mail is the more recently modified directory.
Unfortunately, I can't find man pages for elm, and I've never used it,
so I don't know what the -f switch is supposed to do.
|
