We recommend using MIPCL-PY for rapid development
and writing prototypes, as well as for teaching Mixed Integer Programming.
The only reasonable way of using MIPCL in other languages
for solving complex optimization problems is
- to implement an optimization problem in C++ via a class
derived from CMIP or CLP,
- and then write an interface for using that class
from the target language.
- To implement simple applications, Java programmers can use the JVmipcl,
which is a module that implements
the Java Native Interface (JNI) to MIPCL.
- JVmipcl implements only a part of the MIPCL API,
for details see the JVmipcl
- Complex applications overriding MIPCL virtual functions
--- say, those needed for generating problem specific cuts ---
cannot be implemented by using JVmipcl.
- Efficiency issues are the main reason for such restrictions.
- In addition, debugging a complex program that mix languages is a real headache.
- C# programmers can use MIPCL via the
OPTANO Modeling library.