Talk:Functional logic programming
|WikiProject Computer science||(Rated Stub-class, High-importance)|
|Sources for development of this article may be located at|
Remark on FLP and λProlog
In my opinion, lambda Prolog isn't functional in the same sense as Prolog isn't functional. In both languages there is no definition mechanism for functions.
Functions are only a special case of predicates that happen to be functional in certain arguments that are viewed as input and output. But we want to retain flexibility and to use a predicate in any direction.
In as far lambda Prolog only offers the possibility to define predicates with higher order terms using higher order bodies, and I doubt that this is functional logic programming as somebody would understand it.
- Perhaps you're right — it's a long time ago that I last looked at λProlog. I might have confused HO logic programming and FLP when I wrote this stub. QVVERTYVS (hm?) 17:53, 23 December 2015 (UTC)
- I glossed over the papers of Michael Hanus, one of the codesigners of the Curry programming language. It seems to be that he hardly mentions λProlog. It also occurs to me, that most likely currently the Curry programming language has some important HO features missing, like hypothetical reasoning and quantifiers. Same Problem with the Mercury programming language.
- There is a fine line between adding a constraint store, which both Curry and Mercury did, and providing the HO features that λProlog does. I checked a little bit the papers that introduce a constraint store for Curry and Mercury, but I couldn't find the HO features and I don't know where else to look. Jan Burse (talk) 09:52, 24 December 2015 (UTC)