I don't think there's real difference between Python and C. With Python you can't make your program use more than 4 GB of address space if you run it with a 32-

I don't think there's real difference between Python and C. With Python you can't make your program use more than 4 GB of address space if you run it with a 32-bit interpreter. You have to swap the interpreter. In C the same goes for the compiler. And yes, you can inspect the upper limit by looking at the width of size_t, but it will be seen differently with 32 and 64-bit compilers although the program will be the same. And you _can_ make program behave differently basing on size_t's width, but you're not required to. It doesn't change that fundamentally Python is no more Turing-complete than C just because you can't do it in Python (that's my assumption, I don't know Python well enough actually). Maybe it all boils down to how CPUs work, and maybe it's safe to say that the incompleteness comes from the CPU implementation? You can of course argue that Python interpreters are written in C/C++, but of course we can imagine they can be written in assembly. Edit: after I read some other comments I think I see the point - that indeed the problem is the implementation (on CPU).