[mercury-users] The Logic of Mercury

[Lee Naish]

In message <19990908095111.A25974 at hydra.cs.mu.oz.au>you write:
>On Tue, Sep 07, 1999 at 02:14:01PM EST, Lee Naish wrote:
>> Unless you create data structures containing higher order objects (eg,
>> lists of functions) its easy to optimise away all higher order overheads.
>> This should cover more than 95% of cases.
>
>*should*? Any real data, or just guessing?

Sounds like you are trying to undermine my position without actually
disagreeing with it.  Perhaps you believe it but don't want to.

What percentage of the higher order code you have written uses data
structures containing higher order objects?

I use higher order in Prolog code far more than most people and I have yet
to write any which would be less efficient using apply/3 with a half decent
optimiser.  The only time I can remember using higher order things inside
data structures was the output of my lazy functional -> Prolog
transformation, and that used call/1 so its not an issue.

The only time I've used it in a functional language was a single line
definition just to show students it could be done.

And looking at other people's higher order code I haven't seen it used
either, though I've seen plenty of higher order code.

Sure, there are uses.  I've *thought* of uses a few times but never uses
such things in practice (sometimes its turned out there is a simpler way
which is just as good).

	lee
--------------------------------------------------------------------------
mercury-users mailing list
post:  mercury-users at cs.mu.oz.au
administrative address: owner-mercury-users at cs.mu.oz.au
unsubscribe: Address: mercury-users-request at cs.mu.oz.au Message: unsubscribe
subscribe:   Address: mercury-users-request at cs.mu.oz.au Message: subscribe
--------------------------------------------------------------------------