Thomas wrote:
Karlos wrote:
Like I said, wrong. Evil, infact, :lol:
Well, with a BPTR you could theoretically address up to 16 GB of memory as opposed to the 4GB of a usual C pointer. Of course the processor architecture has to support it.
Bye,
Thomas
Well, first off there's nothing in C that says a pointer is 32 bits. This is a machine implementation. Secondly, the 68K has this particular 32-bit pointer implementation anyway, so the BPTR offers nothing. Lastly, I don't know of any common architectures that do perform this kind of n*sizeof(element) type addressing so in most cases, it's simply a total waste of cycles to store it in this fashion when at best you are going to have to shift it every time you want to dereference it ;-)
As I said. BCPL pointers are wrong, evil and they smell too :lol: