KevinL écrivait : > On Wed, 2002-09-18 at 06:05, Michael Renzmann wrote: > > "killall .bugtraq" would be suitable as well, and it would "destroy" > > every other instance of the program that is running currently. Even if > > detecting the current PPID does not work for whatever reason. > > Solaris is vulnerable to this bug? Solaris "killall" kills _everything_ > - not just the named process. > 'course, given they're theoretically not running the webserver as root, > this shouldn't be a huge issue... But it's not as social as you might > think.
I agree... Even on Linux the killall command had not always given the same result. I can remember some killall on (very) old RedHat that even didn't work properly. And don't forget killall has been a script sometime! So it may require additionals environnement variables (PATH) and so on... But may be the main point is: is it really possible to have multiple instance of the .bugtraq program?!? If so, all of them would join the network and should receive the mail-sleep-kill command! J.C.
