interesting, I'll try that. I usually set it to 6 ultrapeer connections because search results seem much better when I do. However I'm pretty sure I've had the crash when I did not bump up the connections.

I just had the crash again and checked out a couple things (Limewire was in the background and I saw that it had crashed because it wouldn't come to the foreground when I clicked on it's window)

-- the out.txt and err.txt files were empty, so Limewire didn't log anything
-- once again I was unable to do a ps from a shell
-- tried shutting down, but os x froze as soon as it sent the quit event to Limewire

from now on I'm going to grab the pid after I launch it so next time I can try kill -1 when it's frozen in the background

I have not changed the number of ultrapeer connections - LimeWire does not limit the number of connections that are not ultrapeer (in the user interface anywhere). If I have the many connections, most of them are plain Gnutella ones. I also do have the problem then that most of my bandwith on my DSL line is then eaten up by just the queries.

I just updated to MacOS X 10.1.4 and see if the problem still persists.

I can confirm that the problem does still exist in 10.1.4. It hit me last night. I also believed that it was caused by connecting to more than 3 Ultrapeers (I normally use 5 or 6). This problem has been around for at least the last 3 point releases of LimeWire. It is the only program that takes down the whole OS that I use.

My last crash happened when I was connected to only 3 Ultrapeers, was d/l only 2 files, had around 10 queued and was sharing around 500 files. I was moving my mouse to switch applications but never made it. I got the spinning beachball of death and it was all over. This was after running LimeWire for about 10 min. The bug is not caused by any action of the users, it just seems to be a random lock-up of the entire system.

Running:
iBook 500, OS 10.1.4 (also happens in 10.1.3), java 1.3.1 update 1

Oh, and this bug does not occur using Fern so it may be a problem in the GUI code. Also, I am not using hardware accleration in java so maybe this is just a problem for people not uising hardware accleration.

Hi there!

I noticed this crash too! It hurt me for _5_ times today.

I am running 10.1.4 and had the latest LimeWire (2.3 I believe, although what was strange, it had buttons to also search for slashdot news and stuff which are gone after I downloaded a news version)

My machine is a iMac DV 400 Slot Loading with 386 MB ram.

When LimeWire locks up I am first not able to bring it to the foreground by clicking on its window or its window icon in the Dock. (Clicking the window Icon in the dock would just do nothing)

However when I try to click the Dock *Icon* of LimeWire then my computer instantly locks up.

So that seems to be the culprit for now.

Next time I get such a crash I will try to force quit LimeWire either through the gui or through the cli.

Could it be that this lockup somehow either crashes the window server (but I can still movie the mouse) or deadlocks it so that it waits for a specific message from limewire?

Thanks a lot,

cu Martin

Sorry, I forgot this: I am now using LimeWire 2.3.4 and it was running for the hole night, so maybe the problem is gone. But I can´t warrantee it.

cu Martin

I did let it run all night and it locked up again under OS X 10.1.4. This time I had a shell window in the foreground with the kill command for the pid of LimeWire already entered, just not pressed return. Everything appeared to be frozen (gui wise), but the shell windows still worked. I did press return and the CPU load meter immediatly returned to normal levels, but the LimeWire window did not disappear from the screen and the gui was still frozen. Typing "ps ax" on the command line did not run to completion either. I just had to reset the machine.

It apears that LimeWire is now frozen again (10.1.4 LW 2.3.4).

Do any of you have an Idea what do to?

My shell seems to be unable to start (theres a ps in the startup script as it seems), Prozess viewer also.

cu Martin

Ok, I can contribute some more detail now:

- LimeWire has since now only "deadlocked" my mac if it was in the background, if it was idle (not up or downloading anything) and was running for some time (at least half an hour till now).

- LimeWire did not Crash when it was uploading something. <g>

Whenever it dies, I cant bring it to the foreground by clicking on its window (also when its in the dock).

_But_ I can keep working for as long as I dont try to use any command that tells me something about PID´s on my computer or trys to kill LimeWire in any way.

Whenever I do that my computer just locks up. I cant click anything (but I can use the mouse) and I can´t give it any commands via the keyboard. Also there doesn´t seem to be any kind of timeout that I can wait for.

¿¿ Any Ideas?

Developers are you listening? Or should I/we post this somewhere else?

cu Martin

I was having a serious problem with limewire hanging my machine but it appeared to happen when my system tried to go to sleep. Not just a display sleep but the hard drive sleep. I changed my energy savings options to never have the hard drive sleep and since then I have not had any hangs using limewire.

No, it definitely isn´t the Hardrive sleep. Today LimeWire crashed while using it and my hard drive definitely wasn´t asleep.

I also did some tests where I let my harddrive spin down and switched to limeWire then. It worked perfectly and my HardDrive spun up again.

Although I can´t confirm that when LimeWire is in the backround, the Drive spinns down and LimeWire trys to read from it.. it doesn´t lead to a crash.

cu Martin

p.s.: (Sorry for the bad english)