CVSTrac
A Web-Based Bug And Patch-Set Tracking System For CVS

CVSTrac is bug and patch-set tracking system for use with CVS appropriate for use on projects with up to a few hundred developers. CVSTrac is designed for low-ceremony development - it provides mechanisms for tracking changes and bugs without unnecessary restrictions.

Features

• Automatically generates a patch-set log from CVS check-in comments
• User-defined database queries with configurable colors
• Web-based administration of the CVSROOT/passwd file
• Built-in repository browser
• Built-in Wiki
• Very simple setup - runs as CGI or as a stand-alone web server
• Minimal memory, disk and CPU requirements - works on old hardware
• Access permissions configurable separately for each user
• Allows for anonymous users
• Implemented as a single stand-alone executable (written in C)
• Uses a built-in SQL database engine (SQLite) - no external RDBMS required
• Tested under Linux - probably also works on other versions of Unix
• GNU Public License

Status

CVSTrac is self-hosting. For information on the change history and current status of CVSTrac, or to enter new bug reports or new feature requests, visit the CVSTrac server at http://cvs.cvstrac.org/.

Installing

1. Obtain a copy of the executable, either by compiling it yourself or by downloading a precompiled binary. Put the executable in any reasonable "bin" directory such as /usr/bin/ or /usr/local/bin.

2. Choose a directory in which the program should run. The parent directory of your CVS repository is often a good choice. In the sequel, we will assume that the chosen directory is "/home/cvs".

3. Arrange for the CVSTrac executable to be launched either from CGI, inetd, or xinetd.

   CGI	Make a CGI script similar to the following: #!/bin/sh /usr/bin/cvstrac cgi /home/cvs Use permission bits or suEXEC to make sure the CVSTrac binary runs as a user that has read/write access to the CVS repository.
   inetd	Make an entry in the /etc/inetd.conf file similar to the following: 80 stream tcp nowait.1000 cvs /usr/bin/cvstrac cvstrac http /home/cvs The entry above assumes that the user cvs has read/write access to the CVS repository.
   xinetd	Create a file named /etc/xinetd.d/http with the following content: service http { socket_type = stream wait = no user = cvs server = /usr/bin/cvstrac server_args = http /home/cvs } As before, this entry assumes that the user cvs has read/write access to the CVS repository.

4. Initialize the CVSTrac database by logging in as the user that CVSTrac will run as ("cvs" in the examples above) and typing the following command:

   cvstrac init /home/cvs project

   Substitute the name of the project for "project".

5. Use a web browser to access CVSTrac. If launched by CGI, the URL will follow this template:

   http://host/cgi-script-name/project/

   If launched from inetd or xinetd, the URL will look like this:

   http://host/project/

   Log in as user "setup" with password "setup" and make the following changes:

   1. Change the password. For security reasons, you should do this first.

   2. Go to the "Setup" page and enter the root directory of the CVS repository.

   3. Add additional users. Note that you have to do this after entering the CVS repository name so that the CVSROOT/passwd file can be written as each user is entered.

6. Repeat steps 4 and 5 above for each project that you want to manage using CVSTrac.

A precompiled stand-alone binary for Linux (i386) is available as cvstrac-@VERSION@.bin.gz. A tarball of the source code is available as cvstrac-@VERSION@.tar.gz.