[m-dev.] Re: [mercury-users] Determinism

[Ralph Becket]

Mark Wallace, Tuesday, 10 August 2004:
> 
>    That motivation is a LOT less convincing than the motivation for
>    typing!  (Note that in the context of constraints there is a
>    realistic possibility that we can compile and run ANYTHING that is
>    type-correct, so the user never needs to encounter error messages
>    arising from impossible modes)

I disagree, the documentation alone is useful for specifying whether a
predicate can fail or not or whether it is committed choice or not.
And even though non-determinism may be the common case for CLP, it's by
no means the rule.  But determinism annotation is more than just
documentation: it is checked documentation!

The thing to bear in mind is that mode declarations aren't there to
ensure your code is correct (except with respect to uniqueness), but to
provide operational cues to the compiler and the programmer.

-- Ralph
--------------------------------------------------------------------------
mercury-developers mailing list
Post messages to:       mercury-developers at cs.mu.oz.au
Administrative Queries: owner-mercury-developers at cs.mu.oz.au
Subscriptions:          mercury-developers-request at cs.mu.oz.au
--------------------------------------------------------------------------