From Kolab wiki

Note
The sources available in Kolab CVS are not guaranteed to work all of the time. So there is always a high chance that the following will fail. You should have extensive knowledge concerning the Kolab server and the underlying OpenPKG system to get your Kolab CVS server successfully installed.

This section will give an overview on how you can build your own OpenPKG based Kolab release and include your own development if you desire to do so.

Prerequisites

In order to use the Kolab CVS repository and build packages from the current HEAD revision (or one of the active branches) you will need to provide a sound build environment on your base distribution (the distribution hosting the OpenPKG environment). You will definitely need the following elements:

• a working GCC compiler
• autoconf and automake (a note for Debian systems: You should have automake-1.9 installed)
• the tool CVS

Beyond that you will need a Kolab development environment and an account on the OpenPKG network.

Building a release

There is a script available in Kolab CVS that tries to automate the Kolab server release process. As mentioned above the script requires a full Kolab CVS environment. So before running the following commands make sure you installed a working Kolab development environment.

In theory you should be able to move into your Kolab CVS directory (/kolabdev/CVS/server if you followed the CVS setup instructions) and issue the following command:

./cvs-kolab.sh -r

In the unlikely event that the script runs through without problems your new set of source pacakges will end up in /kolab-packages/source-release/RELEASE.

Problems

The approach mentioned above provides plenty of possibilities for failure. It may fail because of problems with Kolab CVS but it may also fail because of dependency or configuration issues of all the new packages that get pulled in from OpenPKG.

Usually the Kolab developers perform this procedure only every few month when they are about to prepare a new release. Around this time the likelihood for failure of the script is significantly reduced. But in between releases there is a high chance that the combination of Kolab CVS changes and upstream changes prevent you from successfully building a release.