Good take on RHEL's SELinux

Andy Oram has an interesting article on his blog about how his approach to RedHat's targetting SELinux policies changed once he really thought about their intended deployment model. I won't give away the ending, but you can read it for yourself.