Dear Denys, Great to see you back here, sorry for my late response.
I only have a brief question concerning your prolog and python bindings: Your prolog bindings doc says "all constraint and branching posting functions are available just like in Gecode". Does that mean branchings (distribution strategies) can be freely defined with these (as in Mozart 1.*)? Thanks! Best wishes, Torsten -- Dr Torsten Anders Course Leader, Music Technology University of Bedfordshire Park Square, Room A315 http://www.torsten-anders.de On 7 Mar 2012, at 20:14, Denys Duchier wrote: > Hi guys, > > I just read: > > https://github.com/yjaradin/mozart2/wiki/Gecode%20Glue > > and I am wondering why you think that garbage collection requires > synchronizing with search engines? > > I am maintaining gecode bindings for both prolog and python, and, in my > opinion, it would be fairly straightforward to use the same approach for > mozart2 bindings. My approach is this: > > 1. use oxygen to extract XML from gecode > 2. use XSLT to extract profiles for constraints > 3. use a code generator to generate bindings for them (in particular, > the multiple dispatch on all arguments) > 4. use a handwritten library to wrap everything in a generic interface > > Cheers, > --Denys > PS: just in case you don't know me: I used to be on the Oz team. _________________________________________________________________________________ mozart-hackers mailing list mozart-hackers@mozart-oz.org http://www.mozart-oz.org/mailman/listinfo/mozart-hackers
