Re: residuation for compositional search?

From: Antonio Regidor García <a_regidor_at_yahoo.es>
Date: Wed, 15 Jun 2011 08:00:53 +0100 (BST)

--- El mié, 15/6/11, Michael Hanus <mh_at_informatik.uni-kiel.de> escribió:

> Whether or not an error message is provided is a question
> of the programming environment. If I want to compute
> something
> and get no result, I consider this as an error, i.e.,
> I have to improve my program. For instance, consider the
> expression (head []). In Haskell/GHC you get, as a service
> of
> the run-time system, an error message. In Curry
> implementations,
> this expressions fails and does not produce an error
> message.
> However, if you turn the debugger on, you might get a
> message
> about this failure.

I mean, in other languages and systems, you have:

infinite looping vs. stopping and warning the user

Here, you have, in the first version of the program:

infinite looping (PACKS) vs. stopping and no warning (MCC)

due to the different evaluation strategy. MCC computes all solutions and prints "No more solutions." There is no clue about something can go wrong with other compilers. And, probably, the debugger can't see anything wrong in the program compiled by MCC. For MCC, the program is correct and terminates gracefully.

Best regards,

Antonio

_______________________________________________
curry mailing list
curry_at_lists.RWTH-Aachen.DE
http://MailMan.RWTH-Aachen.DE/mailman/listinfo/curry
Received on Wed Jun 15 2011 - 18:09:39 CEST

This archive was generated by hypermail 2.3.0 : Mon Nov 11 2019 - 07:15:08 CET