|
|
| Author |
Message |
| 25 new of 234 responses total. |
scg
|
|
response 20 of 234:
| 
Jun 25 04:58 UTC 1998 |
The idea is to find out how many users we can support without bogging the
system down. Bogging the system down is a bad idea, but we don't gain
anything by making people wait in the queue while there are spare CPU cycles.
|
senna
|
|
response 21 of 234:
|
Jun 25 04:59 UTC 1998 |
Heh, I remember when I measured the quickness of pine in portions of hours.
I'd occasionally have to bump a key to prevent myself from getting lagged out
from inactivity while it loaded.
|
mcnally
|
|
response 22 of 234:
|
Jun 25 06:02 UTC 1998 |
One thing to keep in mind is that Grex has had a couple of incidents
recently with people abusing the mail system. Sending out a huge
batch of mail can cause the load to skyrocket very quickly..
|
keesan
|
|
response 23 of 234:
|
Jun 26 14:24 UTC 1998 |
STeve said there was a vandal on the system at the time when Pine was running
slowly, so it has nothing to do with number of users. Pine is back to
amazingly fast again today.
|
srw
|
|
response 24 of 234:
|
Jun 27 05:48 UTC 1998 |
High load averages are almost never a function of user count. Really
high numbers like in the 70s usually indicate a resource hog, which can
be either a deliberate attack or clueless behavior. If a grex staffer
finds the system in that mode, he or she will take steps to restore
normal performance.
|
other
|
|
response 25 of 234:
|
Jun 28 06:58 UTC 1998 |
i just dialed in twice but got nothing after:
CONNECT 19200
actually, twice at 7613000 and once at 5041.
i'm on via telnet now.
|
scott
|
|
response 26 of 234:
|
Jun 28 12:54 UTC 1998 |
Re. 25: The terminal server had crashed. I've rebooted it.
|
kewy
|
|
response 27 of 234:
|
Jun 30 14:32 UTC 1998 |
i'm not sure if this is the right item, but i've been getting severly large
numbers in waiting to telnet to grex, like usually around 40 or 50 in the
morning, any reason why, or just a lot of users in general around lately?
|
rcurl
|
|
response 28 of 234:
|
Jun 30 16:07 UTC 1998 |
I thought you meant a *large* number...but only 40 or 50 - pshaw.
|
ivynymph
|
|
response 29 of 234:
|
Jun 30 22:02 UTC 1998 |
I've been getting either an immedite connection or a number between 12 and
20. I got 33 once, but I just disconnected. THe speed of the connections
has been more pleasant than expected.
|
arthurp
|
|
response 30 of 234:
|
Jul 1 00:00 UTC 1998 |
Seems to me everyone wants to get in from work in the morning. The
queues are quite long then. There's just so many users...
|
tpryan
|
|
response 31 of 234:
|
Jul 1 21:31 UTC 1998 |
Modem server was also unable to do it's thing yesterday afternnon,
in the 5pm hour.
|
aruba
|
|
response 32 of 234:
|
Jul 2 21:31 UTC 1998 |
Whoa, this looks like a record to me:
5:29pm up 5 days, 12:40, 68 users, load average: 107.91, 103.51, 91.55
|
kerryp
|
|
response 33 of 234:
|
Jul 3 01:05 UTC 1998 |
Pipe can't access /usr/lib/sendmail : No such file or directory
This is not a good sign.
|
keesan
|
|
response 34 of 234:
|
Jul 3 14:02 UTC 1998 |
On Thu, 2 Jul 1998, C. Keesan wrote:
(In connection with being told my terminal has been idle for more than
15 minutes and I would be logged off in another five, while in the middle
of a chat with Sreeni.)
> I think I was reading agora and typed !talk. This has happened to me
> before. Should I exit bbs to avoid this problem when chatting? I don't
> know any other prompt besides the ok: in bbs where I can type !talk. I
> get chat requests while using lynx but I think that I have to quit lynx
> before chatting, at least that is what I do.
> I was dialed in, Sreeni telnetted, as was the case last time I had this
> problem.
>
It is likely that you were experiencing a problem in your pager. I
seem to remember that the "less" pager doesn't report idle time. If
you have a shell that supports multitasking (i.e. if you aren't
using the Bourne shell [sh]), you can hit control-z to get back to a
shell prompt, from which you can issue your talk command. Doing If
you do this, you should not be affected by your pager's inability to
accurately report its idle time. When you are done talking, from
the shell prompt you can issue the
fg
command, which stands for "ForeGround," to return to your BBS session.
(I was also told that I can probably type !talk from lynx, but I will try
the Ctl-z and fg instead, as chats can last hours!)
> Sreeni said my write permissions were off another time when he tried to
chat with me but I know they were on.
. I think it is just that I was in Pine and pine must tell people that
> write permissions are off - I also cannot get tels while in pine. So
> Sreeni e-mailed me about my permissions being off, and I then tried to
> chat with him but he was gone.
>
Pine has a setting that automatically turns off your messages while
it is run. You can disable this feature by typing
msc
from within Pine, and by selecting the "Allow Talk" field from the
menu of options.
Please let us know if we can be of additional assistance.
____________________________________________________________________
Michael O'Leary mike@oleary.org
|
keesan
|
|
response 35 of 234:
|
Jul 3 19:15 UTC 1998 |
Another problem, when I tried to chat with someone using the menu shell we
were unable to connect until I went to the menu and typed T for talk from
there. Why?
|
davel
|
|
response 36 of 234:
|
Jul 4 12:10 UTC 1998 |
Possibly because they were trying to use talk & you were trying to use chat?
|
keesan
|
|
response 37 of 234:
|
Jul 4 13:23 UTC 1998 |
Nope, we tried it both ways, each of us typing talk, I think. We ended up
without the split screen that I usually see in talk, and got our lines of
typing scrambled with each other's until we learned to do 'o'. I have
had the no-split-screen mode before when doing talk from the menu. Can
someone explain what is going on and how to tell talk and chat apart and how
you get into them? I prefer the split screen. (By both ways I mean that we
each requested chat, I think - I am not sure what John was doing, it was his
first attempt at online talk).
|
danr
|
|
response 38 of 234:
|
Jul 5 23:41 UTC 1998 |
Dunno if this is a bug or a design enhancement, but Backtalk has
decided to display the 'view previous responses' line up near the
top of the page instead of after the unread responses. If it's a new
feature, I kinda liked it the other way.
|
arthurp
|
|
response 39 of 234:
|
Jul 6 03:31 UTC 1998 |
I found that Backtalk falt refused to talk to IE 2.? for NT4. It just
gave a crash screen that said contact the authors. I didn't feel it was
important enough to report since it only seems to affect IE 2.?.
|
srw
|
|
response 40 of 234:
|
Jul 6 04:49 UTC 1998 |
Dan, the last change to the pistachio interface was in March, as far as
I am aware. Jan might have some comments on the preferred location of
the 'view previous responses' line. I uses the no-frame switch, so I
also get it only at the top. Doesn't bother me much, though.
Charles, I wasn't aware of such a situation with IE2.x such as you
suggest. There are some known issues, but they are not major, as far as
I know. I think I have IE 2.x for some x on my NT machine at work. I
usually avoid using it, in favor of Netscape, but I thought it worked ok
last I tried.
Can you send us some details? backtalk@hvcn.org thanks
|
ejb
|
|
response 41 of 234:
|
Jul 6 11:33 UTC 1998 |
I've seen 400+ load on Grex sometimes..
|
aruba
|
|
response 42 of 234:
|
Jul 6 13:16 UTC 1998 |
Since we went to this machine? I think the previous machine would take an
hour or so to run "uptime" with a load that high.
|
arthurp
|
|
response 43 of 234:
|
Jul 7 02:34 UTC 1998 |
Hm. That was with a project machine at my place, and with a laptop at
my place. The laptop is gone/sold, and the project now has OpenBSD on
it. It'll be a while before I get the chance to try NT on it again. I
can certainly see why you would avoid IE2. It's won't even do anything
while it's downloading a file. (I was using it to download a real
browser both times.)
|
janc
|
|
response 44 of 234:
|
Jul 8 02:56 UTC 1998 |
The "view previous responses" thing has always been at the top of the page.
My rationale was that if I was looking for response 12, i'd probably first
try scrolling back. If I found that the first response was 16, then I'd look
for a button. Since I'm probably at the top of the text then, the button
should be at the top of the page.
As for IE2, I think you should sue Microsoft.
|
