Building wxSA on Linux

From LSDevLinux
Revision as of 12:24, 14 October 2010 by Mayhewn (talk | contribs) (Remove some categorizations)

Jump to: navigation, search

Checking out the code

The Subversion repository is at You can check out the trunk as follows:

$ svn co SpeechAnalyzer

It's also possible to set up a local git repository which mirrors the main Subversion one. You can create your own local branches and make a series of commits to them. Then, when you're ready, you can push a series of commits back to the trunk. Consult the git-svn man page for instructions on how to do this.

Installing necessary packages

First, you will need to install various packages to get yourself a standard C++ development environment. Start with build-essential and work up. Gedit is a very adequate programmer's editor, although you'll probably want an IDE. KDevelop is good for C++, although there are various others if you don't already have a preferred one. Eclipse is the 600 pound gorilla, but is very capable if you have a reasonably recent machine.

Beyond this, SA needs wx libraries and headers. Although it's possible to build a latest-and-greatest wx from source, the system-supplied wx is usually sufficient if you are working on a reasonably recent release of your distro. On Ubuntu 9.04, for example, the following packages can be installed:

  • libwxgtk2.8-0
  • libwxgtk2.8-dev
  • libwxgtk2.8-dbg
  • wx2.8-doc
  • wx2.8-examples
  • wx2.8-headers
  • wx2.8-i18n

SA also uses the libsndfile library, which is found in the following Ubuntu packages:

  • libsndfile1
  • libsndfile1-dev
  • sndfile-programs

SA will soon be using the RtAudio library, but this doesn't have an Ubuntu package and so must be built from source.


There is no configure step, so running make is all that needs to be done. The application can then be launched by executing bin/SAminimal.