Re: request for a init script policy
On Jul 02, Henrique M Holschuh wrote:
> I've had a look on what they have there, and it would require some work to
> implement the stuff in Debian. It goes well beyond making sure all init
> scripts return status 0 on OK ;-) And some of our init scripts ARE a mess.
> Thankfully most of them are not that bad, but still...
>
> It is doable, but is it worth the hassle to require those changes _right
> now_ and require them again later when LSB compliance becomes a goal?
I doubt LSB is going to change the standard much once people start
implementing it. If nothing else, maybe we should look at building a
standard init script template for reusability; i.e., something where
we can do:
File /etc/init.d/sendmail:
/etc/init.d/init-script /usr/sbin/sendmail "sendmail SMTP service" $1
File /etc/init.d/sysklogd:
/etc/init.d/init-script --daemons=2 /sbin/klogd /sbin/syslogd
"system logging daemons" $1
File /etc/init.d/nfs-server:
/etc/init.d/init-script --needs-file /etc/exports --daemons=2
/usr/sbin/rpc.nfsd /usr/sbin/rpc.mountd "NFS servers" $1
Obviously more parameters would probably needed for "reload" to work
(pid file, signal number), but you get the idea. This way, whatever
standards compliance is needed can be handled in the init-script
program.
Chris
--
=============================================================================
| Chris Lawrence | Get rid of Roger Wicker this year! |
| <quango@watervalley.net> | http://www.lordsutch.com/ms-one/ |
| | |
| Grad Student, Pol. Sci. | Are you tired of politics as usual? |
| University of Mississippi | http://www.lp.org/ |
=============================================================================
Reply to: