erm .. yes ... SE Linux does get around the issue, that is true.
And (far from being an SE Linux expert, though i have completed the Red Hat training / certification on that one [alas i don't get to use it enough to become an expert]) i have to say that when a system can get itself into a tight enough knot where the setroubleshoot service can not start because selinux is blocking it, and when it does start it craters anyway. Or when relabelling all files at start up, or using restorecon or fixfiles yields files that are in the same directory and rotates of the same log file, created by the same *syslog daemon i feel that 1) i need to re-read my course notes, and 2) there are still some issues to be resolved.
That said, it worked extremely well during the course ... lord knows i may have screwed something else up.
Also, i don't think that anyone in their right mind, even after running in permissive mode without errors truly expects that they can turn on selinux and not have errors. SE Linux is great ... but it's far from flawless. And the tools to support it are also far from flawless. And the Documentation that ships with the distros ... is somewhat lacking.
... but you know what the hilarious aspect of this is? I disabled selinux yesterday because it kept resulting in crashing my multipathd service. Fingers crossed i find some time monday to sort out the selinux issue so i can re-enable it

But darn do i want to use SElinux ... it would be nice to put that training to use
chakkerz