new version of systemd seems to break autopkgtests of policykit-1, kwin and python-dbusmock

[Paul Gevers]

Hi systemd and kwin maintainers,

On 23-06-18 14:38, Paul Gevers wrote:
>> The breakage in kwin looks unrelated to me (maybe caused by the libinput
>> update, dunno).
> 
> To avoid any doubt, I have triggered a reference run, with only testing
> (results in a hour or so). But the last successful run was on 19 June
> and had libinput 1.10.7-1, which is the same version that systemd was
> tested with. I checked the log of the systemd run, and nothing spurious
> was installed from unstable, only the systemd binaries. On the other
> hand, the test for libinput 1.11.1-1 had the same issue, and also there
> no spurious packages from unstable were installed. So I fear we already
> regressed testing and the systemd maintainers can further ignore the
> kwin failure.

The reference run of kwin in testing passed¹. Really the only delta I
can spot, before and after the build (apparently the order of output can
vary), is the version of systemd's binary packages.

@kwin maintainers, do you have any idea how the libinput run and the
systemd run can fail in seemingly the same way, while it seems that both
runs were done with all binaries from testing (where a pure testing run
passes) and only their own respective binaries from unstable.

Paul

¹ https://ci.debian.net/data/autopkgtest/testing/amd64/k/kwin/498051/log.gz

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 488 bytes
Desc: OpenPGP digital signature
URL: <http://alioth-lists.debian.net/pipermail/pkg-systemd-maintainers/attachments/20180624/77331ca1/attachment-0002.sig>