Hi, I admit it was a very bad idea, .... the rational behind it was to develop a standalone widget that would give an up-to-date indication of the amount of free disk space without having to know about the processes performing disk operations. This is why I used a timer (OK, the polling frequency was very high). Mat On 21 Jun 2004, at 17:03, Justin Walker wrote: On Jun 21, 2004, at 1:44, Matthieu Kopp wrote: Hi, After spending a few more hours on the problem, I finally found what was causing the panic...and more importantly for me, how to avoid it. I am still posting a bug to Apple, because as you mentioned it, this should never happen. Be sure to add this to your bug report :-} What triggers the panic in my application is the following: my application is multi-threaded and as I mentionned, I have one thread saving data at regular intervals (using writeToFile:atomically of NSData). Another thread is triggered by a timer and is checking the amount of free disk space every second using : Maybe I'm unclear on what is going on, but it seems to be a terrific waste of resources to check once per second, on a thread devoted to that task, for freespace. Wouldn't it be better to check before you do the write? In general, polling is A Bad Thing, except in dedicated devices (and then, only when the polling doesn't involve "going external"). There are always exceptions, but you should try to design polling out of your application. Regards, Justin -- Justin C. Walker, Curmudgeon-At-Large * Institute for General Semantics | It's not whether you win or lose... | It's whether *I* win or lose. *-------------------------------------- *-------------------------------* _______________________________________________ darwin-kernel mailing list | darwin-kernel@lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.