John Lloyd wrote:
> However, put yourself in the position of having to explain the computational
> model of Escher or Curry to an (inexperienced, say) programmer...
>
> In fact, this is just the Haskell story except Escher allows variables
> in expressions. Having taught Haskell to our big first class here, I feel
> very confident that I can explain this story and pretty soon have students
> writing conventional Haskell programs as well as LP-like ones.
In fact, I already did it in a course on declarative programming.
I started with Haskell and extended the lazy rewrite model to
allow the necessary instantiation of free variables which leads
naturally to logic programming. For this course, I used, as
you can imagine, Curry. A description of this experiment and
the course will appear in the proceedings of the PLILP'97.
You can find the paper also at
http://www-i2.informatik.rwth-aachen.de/~hanus/publications/papers/DPLE97.html
> This I have discussed earlier. But note one additional point. It isn't
> always disastrous for a computation to not compute far enough (i.e flounder).
> Whatever happens the first and last expressions are equal.
But isn't it disastrous to loop instead of computing a value?
Best regards,
Michael
Received on Do Jul 03 1997 - 19:34:00 CEST