[mercury-users] Pred defns

[Tyson Dowd]

On 30-Mar-1998, Fergus Henderson <fjh at cs.mu.OZ.AU> wrote:

> Well, that's fairly vague, I guess.  I have a precise specification
> (150 lines of code) if anyone's interested.
> 
> Comments?

At a glance, I don't see any particular technical problems with this,
however it should be noted that this still leaves the user doing the
work of threading arguments -- only naming them has been automated.

Some of the other proposals (EDCGs for example) allow naming and
threading to be done automatically.  This means the old example of
adding a state pair to some code is almost completely painless, compared
to this proposal, where it is only somewhat painless.  I believe EDCGs
also allowed errors to be avoiding by handling argument ordering, this
proposal doesn't do that either.

Practically, I can see a lot of errors being made by people putting the
wrong operator in the head, or using current when you mean next.
The error messages for these might be difficult to decipher (if you
even get one, it might just run the code in the opposite direction
instead).

So I guess I'm not convinced this proposal has enough in it to use it
instead of a EDCGs.  I'm also not convinced that there are enough
problems with EDCGs not to use them.  Perhaps I should re-read the EDCG
proposals.

-- 
       Tyson Dowd           # So I asked Sarah: what's the serial number on
                            # your computer? She replied:
     trd at cs.mu.oz.au        #          A-C-2-4-0-V-/-5-0-H-Z
http://www.cs.mu.oz.au/~trd #