fork bomb, instability in 2.8.5, 2.8.6

[AlHaz]

I have a problem where 2.8.5 and .6 - sometimes immediately upon execution but sometimes upon becoming an ultrapeer - spawn hundreds (literally) of java forks and become unusable. This happens with the sun 1.4.0 vm and with the ibm 1.3.1 vm

with 2.8.6, it's worse because it starts popping up error dialogs faster than i can close them. I eventually go to a different console to killall -KILL java. I am unable to copy/paste the contents of the error dialog because xwindows becomes unusable due to constantly shifting focus.

Also in 2.8.6 there is a tendency for the ui thread to stop functioning (no screen update) when i try to look at the connections tab, though it's clear limewire is still running in some sense because it's using plenty of cpu cycles and bandwidth. I've waited as long as 20 minutes for it to come back before killing it.

I don't have any of these problems in 2.7.13, but I've spoken to a friend who has the same problems with 2.8 versions. I believe this is probably the same issue as another poster mentioned, having to do with java out of memory errors. The number of searches and number of downloads don't have anything to do with this. "ps ax | grep java | wc -l" returning 241 does.

It doesn't matter, but my linux is debian/woody running on a two-processor intel system with 640 megs of ram, and i am launching limewire with an unmodified runLime.sh script, and installed using the tarball.