metainit ready for extended testing

[Petter Reinholdtsen]

[Sven Mueller]
> Huh? Why shouldn't the init scripts be able to say: I provide
> portmap/syslog/named functionality?

I'm not quite sure, but I believe it is because it is ment as a level
of abstraction, and not something that normally is fulfilled by a
single script.

Anyway, as far as I can see, the insserv package do not handle virtual
facilities in the provides field very well, so I recommend against its
use.

> If anything, I would have thought mentioning some other daemon's name
> was evil, but not mentioning one of the $* capabilities.

I agree.  I'm not quite sure if this should be considered a bug in
check-initd-order or not, but believe insserv handle multiple provides
fairly well.

I guess we need to read the LSB specification to find the answer to
what the correct behavior is, and if it isn't there, request a
clarification in the LSB standard.

Happy hacking,
-- 
Petter Reinholdtsen