Installing and Setting Up CVS

a part of MarkD's Guide to CVS
The notes here are geared for one-time set up and running CVS on a single computer, not for using the client/server CVS features.

Get The Software

You might already have CVS installed on your machine. Look for a cvs executable in /usr/bin, /usr/local/bin, or in /usr/local/cvs/bin. If you've already got the software, skip down to Set up the Repository

Many Linux distributions come with CVS pre-packaged. If you can do that, all the better. Otherwise, visit cvshome.org and download it. ArsDigita employees can ping the SysAdmin staff. They probably have a CVS tarball lying around they can point you to, and if you ask nicely they might even do the installation.

If you can't sucker someone else into doing it, consult the README and INSTALL documents in the CVS source tarball and follow the directions there to compile and install the binaries. Generally it should just be something like this:

% ./configure
% make
% sudo make install
I've not actually done this myself, but it should work this way.

Set up the Repository

You'll need to construct a repository. At ArsDigita, we use /cvsweb as our default repository. First, make sure there's not already a repository living there. If not, create one by issuing this command:

% cvs -d /cvsweb init

To have CVS install whatever infrastructure it needs.

Important: You'll need make sure the file access permissions are set correctly for the CVS repository. The CVS commands are run as the user ID of the person executing the command, so they'll need read and write access to the repository.

In ArsDigita's case, since everyone is a member of the arsdigit group, this can be done by

You can also get Fancy and set up a seperate CVS group to put folks into to exercise more control over who has access.

markd@badgertronics.com
$Id: install.html,v 1.2 1999/12/23 21:49:09 markd Exp $