Hi Brian
On Fri, Nov 15, 2019 at 01:10:58AM -0500, Brian Gupta wrote:
> Do you think it's ok in any case to remove init scripts. Let's say an
> upstream stops maintaining init scripts,
I would like to know if this is a real world problem. Please provide an
example of a package where the init script actually comes from upstream
and is not maintained in debian/*.init and installed by dh_installinit.
Init scripts are inherent incompatible between distributions. One
designed for RedHat or, worse, SUSE will not really work on a Debian
system, due to differences in used tools. LSB tried to standardize
that, but this effort is dead as well.
I don't know if it is a real world problem. I know of at least one upstream that maintains their own packages for the major distros, and maintained both init scripts and unit files, and as best as I can tell they are now only maintaining unit files, as the distros they actively support are all defaulting to systemd. That said, I don't recall them saying they are explicitly removing support for sysvinit. (It's a RedHat funded upstream that surprisingly makes their own debs for Ubuntu and Debian.)
I guess when I gave that example, it was more of a hypothetical, as I'm not aware of all the reasons that package maintainers would give to remove an init script. I think the more important thing to understand what "designed by upstream to work exclusively with systemd" means, and if we do adopt that wording, make sure everyone judges this the same way (objectively vs subjectively). That said, I'm leaning towards supporting Dmitriy's text with Ian's tweak "and no support for running without systemd is available" (Leaving the door open for the community to step up and provide support for alternative init systems, and perhaps making this distinction less important?)
-Brian
Regards,
Bastian
--
Totally illogical, there was no chance.
-- Spock, "The Galileo Seven", stardate 2822.3