Bug#265801: bootlogd stops too early

[Miquel van Smoorenburg <miquels@cistron.nl>]

On Sun, 22 Aug 2004 03:32:46, Branden Robinson wrote:
> On Mon, Aug 16, 2004 at 01:45:18PM +0200, Miquel van Smoorenburg wrote:
> > 
> > On 2004.08.15 19:06, dean gaudet wrote:
> > > On Sun, 15 Aug 2004, Miquel van Smoorenburg wrote:
> > > 
> > > > I'd say that's a bug in xdm, it should be at S95 or so.
> > > 
> > > there are other S99 scripts as well... on my systems i find:
> > > 
> > > S99fetchmail
> > > S99rmnologin
> > > S99stop-bootlogd
> > > S99xdm
> > > 
> > > it's really only by luck that S99stop-bootlogd is almost last.  that's why
> > > i suggested fixing bootlogd rather than trying to discover all the other
> > > S99s and change them.
> > 
> > But that is what should be done. Policy says that if you use non-standard
> > start/stop levels, you have to discuss that with the sysvinit maintainer.
> > Nobody appears to do that. Bootlogd is part of sysvinit, so it's start/stop
> > level has been discussed with the maintainer, so it is correct. All the
> > others are incorrect.
>
> I asked you *months* ago what sequence number you thought xdm should use,
> and I got no reply.

Really? Then I have to apologize.

The amount of e-mail I get is getting so much out of hand that
I'm sometimes not able to read it all. And when unread mail gets
deeper and deeper into the folder, you know what happens ..

But yes, preferably xdm should be at S95 or so.

However after sarge I intend to propose doing away with sequence
numbers altogether and instead move to a dependency-based
solution. And since bootlogd is also really experimental and
turned off by default, I wouldn't worry about changing the
sequence number of xdm.

Mike.