eggdrop start then nothing

Linux · Post by **Linux** » Thu Nov 10, 2005 6:09 am

you mean to say when you start bot with '-n' it works fine and without '-n' it wont work? (i.e. as mention in the very first post that "without -n it does nothing, no telnet, no connection to irc server")

Larkspeed · Post by **Larkspeed** » Thu Nov 10, 2005 7:34 am

yup thats exactly it

right now my bot is sitting in my channel doing everything it should do but to achieve that I have it running in a second console that I can't close or the bot quits.

without the -n it start' without errors but no connections at all in or out

Alchera · Post by **Alchera** » Thu Nov 10, 2005 11:14 am

I did a test myself today and started a bot with the -n switch, closed the the Telnet session (SecureCRT) and the bot remained connected to the Network and functioned as expected.; in fact it's still connected after 5 hrs.

Your problem, going by my test, therefore has nothing to do with eggdrop in my opinion.

Larkspeed · Post by **Larkspeed** » Thu Nov 10, 2005 11:46 am

ok so you can detach from a telnet session without the session fully terminating. great

but it still does not change the fact that when I run the bot and I Don't use the -n switch which is the correct way to run the bot it does not work.

I don't want it running in a second console with the -n switch on so it is outputing everything to the screen I want it to run in the background like it is supposed to.

I can run it in a screen session with the -n switch and detach from the screen which is effectivly the same thing you are doing just a different way.

However doing it this way I cannot use the botcheck script to make sure the bot is running and restart it if it crashes.

Alchera · Post by **Alchera** » Thu Nov 10, 2005 11:01 pm

As far as I can determine, this has nothing to do with eggdrop. The problem is being caused elsewhere.

Either reinstall tcl8.4 or downgrade to tcl8.3.

Try running the bot with only default scripts (or none).

Larkspeed · Post by **Larkspeed** » Fri Nov 11, 2005 4:57 am

you know what I give up

There realy is no point

if anything other than the bot itself was at fault it would not work at all rather than only not work when it is in the background.

I'll just scrap eggdrop entirly and use a different bot

prim3 · Post by **prim3** » Fri Nov 18, 2005 10:04 pm

Hi,

I installed yesterday a eggdrop and got the next error on connect:

[03:03] === xxxxxxxx: 0 channels, 0 users.
[03:03] main: entering loop
[03:03] Trying server 194.109.129.220:6667
[03:03] * Last context: tclhash.c/238 []
[03:03] * Please REPORT this BUG!
[03:03] * Check doc/BUG-REPORT on how to do so.
[03:03] * Wrote DEBUG
[03:03] * SEGMENT VIOLATION -- CRASHING!
Segmentation fault (core dumped)

anyone an idea what to do?

try'ed multiple server, no result...

demond · Post by **demond** » Sat Nov 19, 2005 12:20 am

as the message says, report this bug (not that someone would bother anyway, they hadn't released a new version for ages and I doubt anyone's working on fixing bugs)

prim3 · Post by **prim3** » Sat Nov 19, 2005 7:45 am

so... what should/could i do then? use a older version then?

demond · Post by **demond** » Sat Nov 19, 2005 1:08 pm

get rid of any scripts you might be using and try again

De Kus · Post by **De Kus** » Sat Dec 31, 2005 1:20 pm

I just helped a sysadmin tracking down the background problem. It seems this because of version missmatching of TCL. Thinking of the nature of the error its no wonder installing and using TCL 8.3 also solves the problem, but thats strictly seen a workaround and no solution.
However I found out that there were 2 TCL versions installed, one in /usr/lib and one in /usr/local/lib. the one in /usr/lib had no header files, version was 8.4.9 and was configured as threaded. In /usr/local/lib there was 8.4.12 configured as non-threaded and with header files. However the configure script detected the TCL in /usr/local/lib, found its headers and compiled. But the system checked the libary first in /usr/lib and ran a non-threaded eggdrop on an old threaded TCL.

Solution A: (the user solution)
use 'export LD_LIBRARY_PATH=/usr/local/lib' to point to the libarary that the bot actually is compiled with. Use the configure --with-tcllib= and --with-tclinc= switchs to change which TCL eggdrop is compiled with.

Solution B: (the root solution)
locate and delete the second TCL installation and preferable install it into the system default prefix (which is /usr and not /usr/local for debian as example). 'whereis tcl' or 'locate tcl8.4' might help you finding duplicates.

PS: no garantue every background connect failure is related to this, but it likely is the main reason. If people can confirm this solution it might be added to the FAQ ^-^.

demond · Post by **demond** » Sat Dec 31, 2005 1:23 pm

how did you know that?

De Kus · Post by **De Kus** » Sat Dec 31, 2005 1:25 pm

demond wrote:how did you know that?

How did I know what?

demond · Post by **demond** » Sat Dec 31, 2005 1:41 pm

that the crash that guy experienced was due to threaded libtcl

btw, there is no such thing threaded eggdrop

De Kus · Post by **De Kus** » Sat Dec 31, 2005 2:05 pm

this is no solution for the crash, but for the original topic which is about eggdrop working in foreground but fails to handle network connections in background. Same problem occured in tclsh so I tried to track it down, since I knew it was sure it must be because of TCL.