Allow slicing multi-pointers as `ptr[start..]`

[dweiller]

As discussed here it may make sense to interpret slicing a multi-pointer p without an end by making p[start..] equivalent to p + start.

[nektro]

related #14611

[likern]

I think this breaks idea of Zig being simple and intuitive. p[start..] implies we want range, but lazy to provide end and saying "Zig use end of p slice", this is intuitive.

But having p[start..] working as p + start overloads this semantic and instead of a range or error indicating we a doing something wrong (slicing multi-pointer without end) we implicitly get weird behaviour (possibly bug).

[N00byEdge]

But having p[start..] working as p + start overloads this semantic and instead of a range or error indicating we a doing something wrong (slicing multi-pointer without end) we implicitly get weird behaviour (possibly bug).

How is that something wrong? That's exactly what multi-pointers are, a slice base pointer.

[likern]

I think I get it wrong, I thought it's about rewriting slice semantics. Yes, slicing pointer this way might be intuitive.