Did *ANYONE* read Martin von Loewis' post about the sysctl(2) interface?
It looks very promising. I'm looking into it to see if it's applicable to a general procfs cleanup. I think others should look at it too.
-Dan