[oe] [meta-oe][RFC 00/27] systemd / initmanager rework

[Andreas Müller]

On Tue, Feb 7, 2012 at 6:14 PM, Koen Kooi <koen at dominion.thruhere.net> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Op 07-02-12 17:24, Andreas Müller schreef:
>> On Tue, Feb 7, 2012 at 4:52 PM, Koen Kooi <koen at dominion.thruhere.net>
>> wrote:
>>> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
>>>
>>> Op 07-02-12 16:12, Andreas Müller schreef:
>>>> The RFC series is intended to
>>>>
>>>> * reduce efforts/pitfalls when implementing native systemd support *
>>>> enable a global configuration for selecting of systemd/syvinit
>>>> (sysvinit needs further efforts)
>>>>
>>>> The whole series' can be found at (branches 'systemd' meta-oe:
>>>> git://gitorious.org/schnitzeltony-oe-meta/meta-openembedded-systemd.git
>>>>
>>>>
> meta-angstrom:
>>>> git://gitorious.org/schnitzeltony-oe-meta/meta-angstrom-systemd.git
>>>>
>>>> Notes: * These are my first python experiences - suggestions welcome.
>>>> * In local.conf (or in distro) the configuration variable
>>>> INIT_MANAGER selects the initmanager to be build into an image. When
>>>> changing the selection, a build from scratch is required.
>>>> INIT_MANAGER currently defaults to systemd (see image.bbclass and
>>>> initmanager.bbclass)
>>>
>>> NAK! The initmanager decision is up the image, not the distro. The
>>> distro can use blacklist.bbclass to blacklist the initmanagers it
>>> doesn't support if needed.
>> It's not possible to let the image decide - e.g for polkit/udisks we
>> would need 2 versions: one configured/linked against systemd and one
>> without.
>
> So split out the libs from systemd ${PN} and voila!
OK - I think you are more involved on systemd. Good luck on finding
out what others might need on systemd-less images...

Another question: any good idea saving us from plastering *-systemd all over?

Andreas