Re: Non-determinism, backtracking and Monads

2003-06-11 Thread Andrew J Bromage
G'day all. On Wed, Jun 11, 2003 at 12:36:30PM +0200, Jerzy Karczmarczuk wrote: > It is possible, instead of implementing the *data backtracking* through lazy > lists, to make lazy "backtrackable" continuations, permitting to redirect > the flow of control to produce something else. The two ways

Re: Non-determinism, backtracking and Monads

2003-06-11 Thread Derek Elkins
On Wed, 11 Jun 2003 09:03:32 +0100 "Simon Peyton-Jones" <[EMAIL PROTECTED]> wrote: > Check out "Embedding Prolog in Haskell", which explores exactly the > topic you discuss. > > http://citeseer.nj.nec.com/272378.html > > Simon and what can be considered a followup to that paper, http://c

Re: Non-determinism, backtracking and Monads

2003-06-11 Thread Iavor Diatchki
hello, you might also want to take a look at the "new monad library" i am working on. it is based on the old one, but does some things slightly different. one of the differences is the "nondeterminism transformer" whose implementation is inbetween lists and continuations. it also resembles resum

Re: Non-determinism, backtracking and Monads

2003-06-11 Thread Jerzy Karczmarczuk
Graham Klyne wrote on the subject of powerset through backtracking: The common thread here is a non-deterministic calculation in which there are several possible solutions for some problem. The goal is to find (a) if there are any solutions, and (b) one, more or all of the solutions. Prolog do

RE: Non-determinism, backtracking and Monads

2003-06-11 Thread Simon Peyton-Jones
To: Jerzy Karczmarczuk; Haskell Cafe | Subject: Non-determinism, backtracking and Monads | | At 11:06 05/06/03 +0200, Jerzy Karczmarczuk wrote: | >I permit myself to observe that your powerset problem (and the restricted | >length problem, i.e. the combinations) is usually solved in Prolog, through

Re: Non-determinism, backtracking and Monads

2003-06-11 Thread Andrew J Bromage
G'day all. On Wed, Jun 11, 2003 at 08:37:48AM +0100, Graham Klyne wrote: > I was thinking some more about this comment of yours, and my own experience > with the ease of using lists to implement prolog-style generators, and > think I come to some better understanding. You might find this amusi

Non-determinism, backtracking and Monads

2003-06-11 Thread Graham Klyne
At 11:06 05/06/03 +0200, Jerzy Karczmarczuk wrote: I permit myself to observe that your powerset problem (and the restricted length problem, i.e. the combinations) is usually solved in Prolog, through backtracking, using reasoning/style which adopts this "individualistic" philosophy. powerset(,)