Re: RFC 272 (v1) Arrays: transpose()

Karl Glazebrook:
> Well if a dimension has N elements then the numbering of the elements
> runs 0...N-1
>
> similarly if the shape has M dimensions, then the numbering of the
> dimensions is 0..M-1
>
> the arguments to reshape should be sizes not last elements (i.e. N's
> not N-1's).
>
> I think this is logically consisteny and is now PDL does it.
>
> How does this sound?
>
I liked the 'anything else would be WRONG' argument better--I might use it
myself sometime ;-)

(It sounds fine.)