Le 21/02/19 à 15:09, Chris Lamb a écrit :
Hi Laurent,
Hello Chris,
It would be nice to have a test for the packages that are shipping a cron script/config file but not a systemd .timer.Replying quickly (whilst this will be fresh in your mind) with a couple of questions: * What are the ramifications of not shipping a .timer? I ask so I can expand this into the tag long description. Bonus points if you could come up with such a text.
For machines that are not running all the time (desktop/laptop), cron is not working well.
In the past the task-desktop and task-laptop were (gnome metapackage still is ATM) pulling anacron to cope with this, but for a few months it's not the case anymore (see: #915370 and #913583).
OTOH, systemd timers are properly handling this situation.So my idea was, we can either force the installation of anacron again (I wonder if we shouldn't do that for buster anyway) or we go forward and we move to use systemd timers instead of cron (which allows proper process tracking just like any other service).
* Is this in Policy (etc.) at all? This is for external references
and it somewhat affects the level of the tag,
Mhh that's right, cron jobs are codified in the policy. I was just proposing to add a check to track the progress of this in case we want to change that in the future.
* Whats the canonical check for this? (A binary package installs
something into any /etc/cron.* yet there is no .timer in the same
package?)