On Mon, Feb 3, 2025 at 8:54 AM Ross Burton via lists.openembedded.org
<[email protected]> wrote:
>
> On 2 Feb 2025, at 21:03, Vyacheslav Yurkov via lists.openembedded.org 
> <[email protected]> wrote:
> > Instead of the python re-implementation build the actual systemctl from
> > the systemd source tree.
>
> Very glad to see this - I tried some time ago but the dependency tree at the 
> time was quite invasive to remove so I gave up.
>

As did I, before I did the shell -> python port. In theory I agree
this is a good thing.

> Have you verified that the final rootfs is identical (or in fact better) with 
> those compared to our reimplementation?
>
> > +DEPENDS = "gperf-native libcap-native util-linux-native 
> > python3-jinja2-native"
>
> I’d hope you can remove some more of those dependencies as we just need 
> systemctl?
>

The getent HOSTTOOLS in the other patch concerns me - that feels like
it could easily be a host contamination problem.

> > +Upstream-Status: Submitted
>
> Submitted where?
>

A nervousness I have with this, is this makes a systemd enabled image
significantly dependent on systemd at build time, and the history of
systemd taking patches which don't match their use cases (which I
don't believe this does) is frankly very poor.

-- 
Alex Kiernan
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#210635): 
https://lists.openembedded.org/g/openembedded-core/message/210635
Mute This Topic: https://lists.openembedded.org/mt/110958135/21656
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub 
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

Reply via email to