[m-dev.] Array modes

[Peter Wang]

On Sat, 25 Oct 2014 18:06:10 +1100, Mark Brown <mark at mercurylang.org> wrote:
> 
> But for Peter's proposal it does make a difference, in particular with
> the meaning of the I =< unique constraint. So what I'm asking is, is
> it true (according to the intent of the reference manual) that
> uniq_array(ground) =< unique?  Yes if 'unique' means only the top node
> is unique, no if it means all nodes are unique.

I understood `unique' to mean unique at every node, by the analogy with
`ground' meaning bound at every node.  Page 59 of dmo's thesis supports
this; at least, I can't read it any other way.

If that's so, I agree that uniq_array(ground) =< unique is false,
and the constraint I =< unique is overly restrictive.

Peter