INSTALL

Installation Instructions
*************************

C++:
  This project uses the Boost.Build system which automatically tracks
  link, include, and other dependencies among any number of libraries and
  projects.  Except for using the (unfinished) CMake build configuration, these
  instructions assume that you have Boost.Build installed on your system.  

  General Install:
    These instructions are for installing this software into a standard layout
    where libraries are installed to  <prefix>/lib
    and headers are installed to      <prefix>/include
    DEFAULT:  <prefix> = /usr/local

    For a simple install, just do:
    
        bjam install
    
    To change the installation path, use
    
        bjam --prefix=/path/to/install/into/ install
    
    For help on other options, execute
    
        bjam --help


  Boost.Build:
    If you use Boost.Build, integrating this library into your own project is
    the most easy.  You simply, include a line like:

        use-project /physical : path/to/physical ;

    where the path/to/physical path can be either relative or absolute.  
    This library has two sub-libraries that are useful and which will
    invoke the proper linking to other libraries such as Boost.Regex.

        /physical//physical : does NOT invoke linking to Boost.Regex
        /physical//runtime  : DOES invoke linking to Boost.Regex

  CMake:
    CMake files also accompany the package and make it easier to use in a
    CMake environment.  To use in a CMake environment, just add appropriate
    lines to your master CMakeLists.txt file, such as

        add_subdirectory(physical)

    You will also have to include the proper line for the include paths, such
    as:
        include_directories(
          ${CMAKE_SOURCE_DIR}/path/to/your/code
          ${CMAKE_SOURCE_DIR}/physical/cxx
        )

    NOTE:  the CMake files are a work in progress and may not work yet.



PYTHON:
  For the python implementation, just make sure that the python directory is
  added to your PYTHONPATH (however you choose do this).  As an example, for
  bash,  you could add the following line to your .bash_profile file:

    export PYTHONPATH=":/PATH/TO/PHYSICAL-PACKAGE/python:"


GNUPLOT:
  The gnuplot 'port' is not the most elegant (i.e. the namespaces are just
  tacked onto the names, such as 'arc_centesimal_minute' instead of
  'arc::centesimal::minute' but it has been handy for me. 


OCTAVE/MATLAB:
  For the octave/matlab version, you only have to run the 'physical.m' script to
  get all of the units/contants loaded into memory as variables.  The namespaces
  are loaded as structs, except that the physical::unit(s) and
  physical::constant(s) prefixes are removed (perhaps this should not be the
  case, but oh well).