Descarwin

README

DESCARWIN is concerned with the hybridization of Descartes (AI planning) and Darwin (Evolutionary Computation) in order to designing satisficing planners.

Here are the steps to build the project:

------------
COMMON STEPS
------------

0) Copy install-dist.cmake files in install.cmake
 * $ROOT/install-dist.cmake -> $ROOT/install.cmake
 * $ROOT/dae/install-dist.cmake -> $ROOT/dae/install.cmake

1) ParadisEO:
 * edit the file $ROOT/dae/install.cmake and define the variable PARADISEO_DIR
 * ensure to compile yourself openmpi statically and define the variable MPI_DIR
 * ensure to compile yourself boost::mpi statically and define the variables BOOST_*

2) [optional] use one of the scripts available to generate the build scripts the project, either:
 * use ./configure.py that will prompt for the configuration
 * mkdir builds/ and run ./make_build_scripts.py that will create all the possible build scripts with different configurations
 * run your own cmake command in the directory of your choice (for example: mkdir build; cd build; cmake .. for the defaults)

3) go to the new directory created (cf. release, debug, ...) and run `make`

4) Location of the generated binaries and libraries:
 * dae binaries are available in the directory dae,
 * cpt-yahsp binaries in cpt-yahsp directory,
 * all the libraries are in lib directory.


 ---------------------------
 AVAILABLE OPTIONS FOR CMAKE
 ---------------------------

You can launch cmake with these options in order to customize your version of dae :
"-DCMAKE_BUILD_TYPE=Debug" : compilation in debug mode. Without this option, release mode is default.
"-DDAE_WITH_MPI=yes" : compilation with OpenMPI and Boost::MPI  (message passing interface for parallelization).
"-DDAE_WITH_OMP=yes" : compilation with OpenMP (shared memory solution for parallelization).
"-DCPT_WITH_OMP=yes" : compilation of cpt-yahsp with OpenMP (shared memory solution for parallelization).
"-DCPT_WITH_MPI=yes" : compilation of cpt-yahsp with OpenMPI (message passing interface for parallelization).
"-DBUILD_DAE_MO=yes" : build the Multi-Objective version of DAE