> On 16 September 2015 at 09:31, Kamil Paral <kparal@xxxxxxxxxx> wrote: > > Actually, this was supposed to be fixed (PK was supposed to refresh > > metadata when it detected altered rpm database), but apparently it isn't. > > Of course, the limitation is that packagekitd has to be running at the > time. I don't know if that was the case here. > > Richard. Hey Richard, would it be possible to remember the last rpmdb timestamp when packagekitd starts, stops or when it detects a change, and then compare it when packagekit is started again? That would cover these cases. But I believe I've seen this issue as well a few times recently, and it was during standard system operations (i.e. PK should be always running), so I guess there's still some issue with it. I haven't had time to report it and attach full logs, unfortunately. -- test mailing list test@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe: https://admin.fedoraproject.org/mailman/listinfo/test