Installation from source

1. Build and install supported versions of libxml2 and PVM3. If you don't have root access to your machine, you can install them in your home directory. Be sure to set PVM_ROOT and PVM_ARCH correctly in the latter case. See installation instructions included with libxml2 and PVM3 distributions for more information.

2. Download and untar Tablix distribution into your home directory.

   $ wget http://www.tablix.org/releases/stable/tablix2-0.2.2.tar.gz
   $ tar -xzf tablix2-0.2.2.tar.gz
   $ cd tablix2-0.2.2
   

3. Configure distribution. Check ./configure --help for available options. Defaults should work in most cases.

   $ ./configure
   

   If you don't have root privileges, you can specify --prefix to install Tablix into a subdirectory in your home directory.

   Warning

   If the configure script can't find your PVM3 installation or the --without-pvm3 flag was used, Tablix will be compiled in debug mode and a warning will be printed at the end of the configure process. Tablix compiled in debug mode uses a linear genetic algorithm instead of a coarse grained parallel genetic algorithm. This makes debugging easier but also severely decreases the probability that Tablix will find an optimal solution to the given timetabling problem. Solving even moderately complex problems is impossible in this mode. Unless you are debugging the kernel or one of the modules you should always use PVM3 (even if you are going to use Tablix on a single machine).

4. Compile. You can add some optimization flags by setting CFLAGS environment variable. Defining you CPU architecture with a parameter like -march=athlon can improve performance. See documentation for the version of C compiler you are using for more information.

   $ make CFLAGS=-march=pentium
   

5. Install compiled binaries. If you specified a proper --prefix option above, you may not need to use root privileges.

   $ su
   # make install