LSB-compliant init-scripts as release goal.

[Petter Reinholdtsen]

[Gustavo Franco]
> Since Carlos' work is going well and considering that SELinux support
> is still a "pet release goal"[0] and i don't think it will make the
> release, we've a chance to push LSB-compliant init-scripts as a
> release goal now.

I'm not sure what SELinux got to do with this, but will chip in a few
comments.

It would be nice if all init.d scripts have LSB-compliant headers in
time for Etch, but it is not required to handle dependency based
reordering of the scripts, thanks to the way insserv handle override
files (it include dependency info for the init.d scripts missing those
internally).  And it is unlikely that we manage to implement an
update-rc.d which uses this dependency info in time for Etch, and thus
I believe it is better to introduce the dependency info without any
hard deadline.  When Carlos get lintian to report missing headers,
developers will slowly fix the init.d scripts by themselves, and I
hope all packages will be fixed in time for etch+1.

There are other changes to the boot which will improve the boot times
much more than parallelization, and I recommend we focus on those for
Etch instead.

This is not to say that we should stop reporting bugs about missing
headers, nor that we should not ask the release team on their opinion.
After all, adding such headers introduces a very low risk to the
packages, and make it a lot easier to detect errors in the boot
sequence.  And such errors in the boot sequence should be fixed before
Etch. :)

As for how we are on schedule for releasing Etch, I believe we are on
track and will make the release in November/December this year, but
there is a lot of hard work left to do, and we should do our best to
reduce the amount of remaining work.  Adding new release goals is
perhaps not the best way to do it, but on the other hand we need to
verify the boot order before Etch freezes.

Friendly,
-- 
Petter Reinholdtsen