ORIGINAL: Marshall Ellis
The neg memory numbers are returns from me looking at the free mem by directly interrogating the windows API??? Don't understand how neg results could happen. I could try to throw an ignore if neg statement in 1.05 that may allow it to run???
How frequently does it fail?
Negative results can happen if a 32-bit application asks a 64-bit processor for an address range, using an API that has the same name. The application believes it is a 32-bit API, but the processor assumes it's a 64-bit one, and responds appropriately (for itself). But, the number is far larger than any number directly addressible in 32-bits (4 GB), so the leftmost binary digit (the 32rd digit, if you will) is set to 1. Thus, the whole number is taken as a negative number.
My wife used to roll the high score table on tetris due to this bug, but it was only 16 bit numbers at that time (when she hit a score of 32768+1 or higher, her score would turn negative and start back "down" to zero.
And, yes, she found out what happens when you get back to zero again, too: It starts counting up again, but now the high score table was totally messed up, even as far as having names were not spelled correctly. One of the original "buffer overflows".
At LAST! The greatest campaign board game of all time is finally available for the PC. Can my old heart stand the strain?