2016-03-21 11:59 GMT+01:00 Michael Meskes <mes...@debian.org>: > > > > > It may be only serious, but it definitely is not fit for a > > > > > release. > > > > Then systemd 229 is not fit for release. > > > Indeed you're right imo. > > This fix is not even list in upstream v230 changelog > > not really serious. > > This surprised me. Doesn't this bug mean that *all* timers are > potentially broken? If so, I'd say it warrants a mentioning in the > changelog to say the least.
It's hard to reproduce, some works some don't; it's about a variable that is not reseted in a loop; so every setup is different. > Anyway, with systemd-cron replacing cron Hum, it can replace cron, but only on explicit user request. All the cron job consumers depends on "cron | cron-daemon" and will always pull vixie cron first. Moreover, only 0,15% of systemd users also decided to use systemd-cron. What is more worrying is the users of custom timers for who there's no popcon metric available. > It seems to be paramount to me that we get a fixed version > of systemd in as it breaks systemd-cron pretty badly. I guess Martin want it fixed in Ubuntu 16.04; this is also tracked on Launchpad so pending fix will be uploaded really soon. https://anonscm.debian.org/cgit/pkg-systemd/systemd.git Alexandre