Bug#727708: systemd jessie -> jessie+1 upgrade problems

[Russ Allbery <rra@debian.org>]

Adrian Bunk <bunk@stusta.de> writes:

> I was misreading that as referring to the headaches udev had caused in 
> the past for Debian upgrades, not that the systemd udev might be the 
> cause of future upgrade headaches.

> But I do not buy this "We already switched to systemd as upstream of 
> udev, so we also have swallow the rest of it":

I don't think anyone is making that argument.

However, there is a valid related argument that, since we're already using
much of systemd, our integrations will be easier if we also use systemd as
the init system.  I don't think anyone considers that single argument
alone to be conclusive, but it's relevant.

Note that one can make, and people have made, an argument that we should
intentionally make our integrations harder if by doing so we can weaken
coupling between subsystems that we feel should be unrelated.  I'm not
intending to comment on that argument in this message, just to note that
it isn't a counterargument to the point above, but rather an argument
about relative priority.  It's saying that a different goal -- weak
coupling -- is more important than reduced Debian integration workload.

> When not using systemd as pid 1, that risk would be confined to the
> parts of systemd Debian would be using (currently only udev).

There appears to be near-unanimous agreement that Debian will also be
using logind in the near future.

I think "only" in this context is misleading, given that the components
we're going to be using anyway (including kdbus in this; if it's included
in the kernel, I highly doubt Debian will refuse to use it in the long
run) are pretty much the same components that people are concerned about
being unstable.

-- 
Russ Allbery (rra@debian.org)               <http://www.eyrie.org/~eagle/>