[m-rev.] for review: Clarify behaviour of upper bounds functions for array types.

[Peter Wang]

On Wed, 29 Nov 2017 18:35:38 -0500 (EST), Julien Fischer <jfischer at opturion.com> wrote:
> 
> > It might be viable still to make array.least_index and array.greatest_index
> > fail for empty arrays. The names suggest to me that they should always
> > return valid indices on success, more so than min/max.
> 
> I would add, det_least_index and semidet_least_index which throw an
> exception and fail for an empty array respectively.  The existing
> least_index should be deprecated and after a while semidet_least_index
> should become least_index.
> 
> Ditto for gratest_index.

Never seen least_index/greatest_index ever used, to be honest.

Peter