Bug 11779 – pointer to slice is not an input range

As a rule thumb, when the type "S" is an input range, than so is "S*". Indeed, all of "S's primitives" can also be called using a pointer, EG: S* p; p.popFront(); //OK! This works for *all* input ranges... except for slices. This is because the "slice range primitives" are actually non-member functions: int[]* p; p.popFront(); //DERP! Error: template std.array.popFront does not match any function template declaration I think, to be consistent with the "emulate a member primitive", that the range functions in std.array should also work on pointer to slice.

I don't think this can work with random access ranges. Pointers already define opIndex, so it can't be forwarded to pointee range. I think the opposite of this request should be implemented: pointers should specifically *not* be ranges, ever. (as an aside, I dislike how D uses . for pointer member access instead of ->, it's a massive hack in my opinion, and really unnecessary since pointers are rarely used in D).

(In reply to Peter Alexander from comment #1) > I think the opposite of this request should be implemented: pointers should > specifically *not* be ranges, ever. Yes, that would be better.