Hi Ben, thanks for your detailed report. Unfortunately it doesn't crash here... not under linux, OSX or Windows. Can you tell me the stats of your system? like OS, CPU, compiler, python version etc. Also, the error "arithmetic exception" reported in the crashlogs is quite strange, because there's not much arithmetic done in the code. The numbers treated are also clean (no NANs etc.). I'll send you some modified code to find out the exact position of the exception.
thanks, Thomas
B. Bogart wrote:
Hi Thomas,
I hope vacation did treat you well.
So I've cvs updated to flex 0.5.1 in CVS and the newest py (as of today) .
It is certainly behaving differently, but I am also still getting the same crash just at a different time.
I've attached a "gem.pd" and a "simple.pd" only with the gem example will you actually be able to see anything, but the simple example will not depend on anything. of course my py script is also attached.
So now rather than crashing on the creation arguments PD now crashes on the "change" message. Actually it also crashes on "reload." to reload the script. So loading the script the first time does not crash (when opening the patch) but it does crash when reloading the script. "bang" and "reset" messages do not crash pd.
Note the below traces were taken from the simple test patch.
I've also changed my classname "array" to curve3d_array to avoid the possible numarray conflict. Again I've had no crashes on OSX.
Sorry for the pasting of the patches/scripts I'm using X forwarding and the patches are not on my email machine!
B.