Project Benzene README ---------------------- Obtaining the source ---------------------- Checkout the source from sourceforge using git: git clone git://benzene.git.sourceforge.net/gitroot/benzene/benzene benzene ---------- Building ---------- You will need the fuego includes and library objects. To get them, checkout fuego from sourceforge.net. For example, from within your benzene directory: % cd .. % svn co https://fuego.svn.sourceforge.net/svnroot/fuego/trunk fuego % cd fuego % autoreconf -i % ./configure --enable-assert (NOTE: This current build of Benzene is tested with revision 1371 (circa January 2011). It is possible that in the future the trunk version of fuego can differ drastically from this version. It may be necessary to checkout a branch in the fuego repository created near revision 1371 to get the compile to work, or to checkout revision 1371 itself with the "-r 1371" option to "svn co".) Include the "--enable-assert" if you will be doing development on Benzene. The compilation will fail if you want to use assertions in Benzene but not in Fuego. If you will not be doing development or testing, or would prefer maximum speed and performance, then you can safely exclude the "--enable-assert" option. % make Change back to the benzene directory. % cd ../benzene If you checked out your benzene code from a git repository you will need to run autotools to create your configure script. If you downloaded a distribution tarball, go straight to the 'configure' step. % autoreconf -i This should complete without error. Now run configure (skip to here if you downloaded a benzene tarball): % ./configure --with-fuego-root=[absolute path to fuego] --enable-assert=yes "--with-fuego-root" simply tells benzene where it can find the fuego files it needs. Do not use a relative path as this will cause an error in the next step, ie, do not use "--with-fuego-root=../fuego/". Recall that if you compiled Fuego without assertions then you cannot use assertions in Benzene, and so you should leave off the assert option to 'configure'. If you want to support boardsizes up to 13x13, include the option "--enable-upto13x13=yes". This will degrade performance, so only use this if you want to play on 13x13. If you want to support boardsizes up to 14x14, include the option "--enable-upto14x14=yes". This will degrade performance a little more (and will override the 13x13 option), so only use this if you want to play on 14x14. If you want to support boardsizes up to 19x19, include the option "--enable-upto19x19=yes". This will degrade performance even further, so again, only use this if you want to play on very large boards. This option supercedes "--enable-upto13x13" and "--enable-upto14x14", so if you use "--enable-upto19x19" only it will have any effect. 'configure' should finish without error. If 'configure' fails to find BOOST or BerkeleyDB, you may need to direct it to those files explicitly (good luck, buddy!). If all is well, then do: % make This will build everything. If you have a multi-core processor, do "make -j#" where # is the number of cores you wish to use. This will greatly speed up the build process. MoHex and Wolve executables can be found under src/mohex/ and src/wolve respectively. To build and run the unit tests, do: % make check ------------- Development ------------- If you are only modifying existing source files, then a simple 'make' is enough to build the new code properly. If you wish to add or remove source files, then you will have to edit the proper 'Makefile.am' files. If you modify any 'Makefile.am', you will need to recreate the makefiles with 'config.status': [add some source files, make changes in all relevant Makefile.am] % autoreconf -i % ./config.status [this will use the settings from the last './configure'] % make ------------ Installing ------------ There is no need to install benzene. If you do wish to install it, the code will be installed to "${prefix}/bin" and the data files from "share/" will be stored under "${prefix}/benzene/share/". Two locations are checked when loading these data files when the program starts up. The first is the "share/" directory where benzene was compiled. The second is the installed data directory mentioned above. The source directory is checked first, then the data directory. If neither exists the program will abort.