Aha!! That was it! After Tim's encouragement, I tried Spambayes again with the data execution prohibition disabled, and it worked.
Maybe some of the other reported crashes have a similar origin. But my own experience suggests that other Windows apps have similar problems with the option, which I suspect will lead most people not to use it. Maybe other manufacturers won't adopt it anyway. Alas, I dug myself another hole while working on Spambayes last night. I'll seek help for that one in a separate message. Bob > -----Original Message----- > From: Tim Peters [mailto:[EMAIL PROTECTED] > Sent: Tuesday, January 17, 2006 12:55 PM > To: Coe, Bob > Cc: [email protected] > Subject: Re: [Spambayes] Outlook plugin crash > > > [Coe, Bob] > > Can someone with a knowledge of the plugin code tell me whether it > > violates the modern fad of scrupulously avoiding the execution of > > data? Old-time coders like me used to do it all the time. (You can get > > away with it even in recursive and reentrant programs if you're > > careful.) But it's considered a security risk today, and the GX620 has > > a factory-preset BIOS option to prevent it. During the weekend, after > > I sent my original note, I found a program that wouldn't run on my > > computer until I disabled that option. I'm wondering now whether I > > ought to make another run at Spambayes with the option turned off. Any > > advice for me? > > Python never executes data, but _something_ invoked when > SpamBayes runs does. There are long discussions of this on > the SB bug tracker, but SourceForge appears to be hosed at > the moment so I can't get a reference for you now. Look for > bug 988095. > > It's definitely worth trying SB with the option disabled. > _______________________________________________ [email protected] http://mail.python.org/mailman/listinfo/spambayes Check the FAQ before asking: http://spambayes.sf.net/faq.html
