4 ms·
I studied prolog years ago at university and more recently rediscovered the power of solvers (like SCIP). Is not really thought about how the 2 crossed over bef
by aidos 7y ago
I studied prolog years ago at university and more recently rediscovered the power of solvers (like SCIP). Is not really thought about how the 2 crossed over before. How does prolog compare to a solver these days?
- triska 7y agoThe primary attraction of CLP, i.e., the combination of constraint programming (CP) and logic programming (LP), is that you get a full-blown programming language (most frequently: Prolog) and in addition seamlessly integrated solvers (over integers, rational numbers, Herbrand terms etc.). CP can of course be hosted on — or provided as library for — other languages too. However, it blends in most seamlessly with a logic programming language like Prolog as its host language. Regarding efficiency, the top-of-the-line Prolog systems provide many of the most efficient and most sophisticated algorithms for combinatorial optimization tasks that are commercially used. See for example the customers page of SICStus Prolog for several interesting applications: https://sicstus.sics.se/customers.html https://sicstus.sics.se/customers.html
- nmadden 7y agoCLP shares some similarity with Satisfiability Modulo Theories (SMT). Is there much overlap in how they are used in practice? When would you reach for SMT over CLP?