Hi guys,

There is a new initscripts release.

Please test and signoff.

****

Mainly bugfix release

Changes:

* rename /sbin/rc to /sbin/rc.d to avoid PATH conflict
* drop very old deprecated setting of uevent_helper
* mount /dev/pts and /dev/shm with standard options in case they are
not in fstab

Features:

* use bootlogd to record the output from boot
* make it possible to symlink /var/{run,lock} to /run/{,lock} (this
is not done automatically)

Workarounds:

* symlink the old /var/run/daemons to /run/daemons (this should not
be necessary but broken rc.d scripts rely on it)
* reinstate the copying of persistent udev rules from /run to /etc
(this will go away one day due to upstream changes)

On 11 May 2011 20:20, Tom Gundersen <teg@jklm.no> wrote:
> ** symlink the old /var/run/daemons to /run/daemons (this should not
> be necessary but broken rc.d scripts rely on it)

The symlink ends up being at /var/run/daemons/daemons.

On Wed, May 11, 2011 at 10:24 PM, Evangelos Foutras <foutrelis@gmail.com> wrote:
> On 11 May 2011 20:20, Tom Gundersen <teg@jklm.no> wrote:
>> ** symlink the old /var/run/daemons to /run/daemons (this should not
>> be necessary but broken rc.d scripts rely on it)
>
> The symlink ends up being at /var/run/daemons/daemons.
>

Right you are. I'll wait a bit before I push a new version in case
there are other things I need to fix as well.

Thanks for testing!

Cheers,

Tom

[2011-05-12 00:09:04 +0200] Tom Gundersen:
> I'll wait a bit before I push a new version in case
> there are other things I need to fix as well.

Minor thing: the help of the newly renamed rc.d script still refers to
the script as just "rc". I guess $0 should probably be used instead.

--
Gaetan

On Thu, May 12, 2011 at 12:52 AM, Gaetan Bisson <bisson@archlinux.org> wrote:
> [2011-05-12 00:09:04 +0200] Tom Gundersen:
>> I'll wait a bit before I push a new version in case
>> there are other things I need to fix as well.
>
> Minor thing: the help of the newly renamed rc.d script still refers to
> the script as just "rc". I guess $0 should probably be used instead.

I have a pending patch from Seblu to fix this, it got dropped
accidentally. Will include it in the next update (probably tomorrow).

Cheers,

Tom