Last post by: arakula (24 Apr 22:52): Your way gives false results for negative values - shifting the sign bit is not the way to go, you have to replicate it through all of the upper 8 bits. Your method would convert 0xffffff (i.e., -1) to 0x807fffff, which is -2139095041. I hope you'll agree with me that this is not the intended behavi [more...]
↧