[Openmcl-devel] New 0.14 binaries

Raffael Cavallaro raffaelcavallaro at mac.com
Sat Dec 20 16:14:42 UTC 2003


On Dec 20, 2003, at 10:49 AM, Gary Byers wrote:

> If you're seeing the same thing that I'm seeing, the console log
> will show about 150 break loop messagee to the effect that
> "a dead Mac pointer isn't a MACPTR, while executing
> CCL::%ASSOCIATE-JMP-BUF-WITH-CATCH-FRAME".  That certainly doesn't
> sound like an esoteric caching problem; it -does- sound like a
> mundane "how-did-I-screw-this-up-this-time ?" problem.  I suspect
> that there are any number of answers to that question.

Yes, that's what I'm seeing here.

> The caching that I'm suspicious of is done by LaunchServices; the
> cache in question is maintained in a file in /Library/Caches (the file
> name seems to vary between Panther and Jaguar A lot of the information
> that the OS uses when launching a bundled application comes from that
> cache; the Info.plist file and other sources of information aren't
> consulted of that cached information appears to be valid.

After deleting everything in /Library/Caches and rebooting, the problem 
persists, so it probably isn't related to the caching issues seen in 
the past.

You're probably right that resolving the:

 > Error: value #<A Dead Mac Pointer> is not of the expected type MACPTR.
 > While executing: CCL::%ASSOCIATE-JMP-BUF-WITH-CATCH-FRAME
 > Type :POP to abort.
Type :? for other options.

issue is probably the first order of business.

thanks again for all your work,

warmest regards,

Raf




More information about the Openmcl-devel mailing list