[Pythonmac-SIG] MacPython 2.1.1c2 available

Russell E Owen owen@astro.washington.edu
Thu, 26 Jul 2001 15:33:45 -0700


I haven't seen this, at least yet. I ran my earlier tests at least 
partly in the background, but I didn't switch out right away. So I 
had another go at it...

I repeated the test, as follows (G4 466, MacOS 9.1. MacPython 2.1.1c2 
with default memory partitions):
- run ConfigurePythonCarbon
- run PythonInterpreter
- import test.autotest
- switch to Eudora as soon as I saw "test_grammar
it ran fine. 106 tests OK. 3 tests failed: test_longexp test_sha test_zlib.

Good luck tracking this down. I hope it's not one of those nebulous 
"corrupted prefs files causing random damage" type things (well, it'd 
be nice if it wasn't Python, but it'd also be nice to have an 
explanation and be able to do something about it).

Macs certainly aren't the most stable machines around. I get 
occasional crashes -- often in Finder when using cmd-delete to trash 
files or cmd-D to duplicate them. It was even worse when I left my 
SanDick compact flash reader card in. Heck, I can't even reliably 
boot with the shift key down -- a problem I suspect may be due to my 
Apple-supplied SCSI card. One could make a whole religion around this 
stuff ("pray to the god ROM and he will send you system stability").

I hope Mac OS X is more reliable (I'm not quite ready to switch), but 
it sure seems like features trump stability at Apple and Micro$oft.

Regards,

-- Russell

>Running the test on CarbonPython 2.1.1c2 on a G4, 540 MHz, OS 9.1, 
>using the interpreter (not the IDE) I'm getting _really_ odd 
>behaviour.  It runs fine unless I switch another application to 
>being the frontmost one.  When I do, the Python script crashes, 
>sometimes with a KeyboardInterrupt exception, and sometimes with 
>_no_ explanation about the exception raised...