> Okay, that's bizarre. I'm going to try and track this down and see > what I can do about that.
Before you get too far deep in it, I think David identified the problem in a later email. Basically, he has gcc 2.95 for Mac OS X. Unfortunatel,y, it doesn't seem to support computed goto. The Configure.pl was supposed to check if they correctly support cgoto, and disable it if they don't. Given his Configure.pl output, his compiler is failing to compile testcomputedgoto.c, but yet it is still choosing to use the CGoto core when compiling Parrot. The fact that this didn't exist in 0.0.4 would explain why that version works for him. I must admit to not having a full understanding of the configure whole configure process, and how to detect compile failures of stuff like this. Hopefully someone else with more experience can figure out why the cgoto detection isn't working for David? Thanks, Mike Lambert
