OOF2: The Manual

Chapter 1. Getting Started

1.1. Acquiring and Installing OOF2

OOF2 is available from the NIST CTCMS website at

After downloading the program, follow the installation instructions in the README file.

1.1.1. Requirements

OOF2 will run on any Unix system with an X11 server. At the current time, it will not run on Windows. It has been compiled and run at NIST on Debian Linux, Ubuntu Linux, and Macintosh OS X systems.

OOF2 requires a number of other freely available software programs and libraries to run.[1] These must be installed before you install OOF2. To compile OOF2 from sources, you will also require the header files (includes) associated with these programs and libraries. These are usually available as part of a development version of the library software.

Detailed instructions for installing the OOF2 dependencies on a number of operating systems are available at

Table 1.1. OOF2 Requirements

Python

2.4 through 2.7

http://www.python.org

Magick++

 

http://www.imagemagick.org/Magick++/

gtk+

2.6 or later

http://www.gtk.org/download.html/

libgnomecanvas

2.6 or later

http://directory.fsf.org/graphics/misc/libgnomecanvas.html

pygtk-2.0

2.6 or later

http://www.pygtk.org/downloads.html

swig

1.1 build 883[a]

http://www.swig.org/download.html

[a] We are using an old version of swig because it was the stable version at the time we began OOF2. Earlier versions, notably 1.1 patch 5, will not work properly. However, swig is only used to generate the glue code between the C++ and Python portions of OOF2. If you're not going to be developing your own C++ modules, you can use the --skip-swig option when building OOF2, and you won't have to install swig. See the README for the details.


Macintosh users should install Apple's X11.app or another X11 server. X11.app is available from XQuartz or as an option during the OS X installation process.

OOF2 also requires the ability to run the blas and lapack linear algebra subroutines. These are provided with many systems. In particular, on Macintosh OS X they are built into the Accelerate framework, and no special libraries are required. On Linux and commercial Unix systems, they may have to be installed.



[1] We hope to soon have installer packages for popular open-source distributions to simplify this process.