nepero.net

The most important problem in this kind of project is represented by binary
files (tipically images, video, zip files): in fact, CVS work very well with
plain text files (html, php, txt, ...), but there're some problem when
you try to import and manage binary files (for example, a binary file can
contain the string "$Id" and, when we checkout this file, CVS replace them with
file's id).
To avoid this problem:

1. Make a copy of CVSROOT configuration files (the most important thing!)
2. Execute
   root@localhost:workspace # cvs co CVSROOT
3. Edit the file CVSROOT/cvswrappers to tell CVS how to handle each file
   extension (remember that the definition if case sensitive). For example:
   *.zip -k 'b' -m 'COPY'
   *.gif -k 'b' -m 'COPY'
   *.png -k 'b' -m 'COPY'
   *.jpg -k 'b' -m 'COPY'
   *.ZIP -k 'b' -m 'COPY'
   *.GIF -k 'b' -m 'COPY'
   *.PNG -k 'b' -m 'COPY'
   *.JPG -k 'b' -m 'COPY'
4. Finally, commit changes:
   root@localhost:workspace # cvs commit -m \
   > "Added support for binary files"
   cvs commit: Examining .
   Checking in cvswrappers;
   /usr/local/cvsroot/CVSROOT/cvswrappers,v
   new revision: 1.2; previous revision: 1.1
   done
   cvs server: Rebuilding administrative file database
   root@localhost:workspace #

Source: Sean Dreilinger, CVS Version Control for Web Site Projects, 1998
(html,
pdf)]]>