Re: pointer elimination in C

Robin Popplestone <pop@dcs.gla.ac.uk> writes:
>C was, as they would say in Congress, -very unique- in providing pointer
>arithmetic. This simply reflected the fact that good compilers had to be
>shoehorned into a tiny computer, the DEC-11, and there was a simple
>one-to-one mapping between something like *x++ and one machine
>instruction.


I don't believe this explanation holds water. The constructs in question
date back to C's predecessor, B, which did not run on the PDP-11 and was
not compiled.


The pointer arithmetic is directly inherited from BCPL, which had only one
datatype -- the word -- and consequently provided pointer arithmetic as a
trivial side effect of providing integer arithmetic. Indeed, in the early
C compilers the distinction between pointers and integers got pretty fuzzy
at times, although unlike BCPL it was not absent entirely.


The *x++ constructs, according to Dennis Ritchie, were Ken Thompson's
inventions, probably vaguely inspired by the autoincrement/decrement
hardware of the PDP-7 but significantly generalized from them. (And in
any case, the PDP-7 hardware facilities were not used in implementing
them.)


These were deliberate design decisions, not accidents of implementation
constraints.
--
Henry Spencer @ U of Toronto Zoology, henry@zoo.toronto.edu utzoo!henry
--

Return to the comp.compilers page.
Search the comp.compilers archives again.