I demand that Udo Richter may or may not have written... > Darren Salt wrote: >> It also needs to cope with being shut down by a signal without any >> possibility of user intervention. Two levels should be available: stop >> now, and stop when inactive. >> This helps init scripts when upgrading a packaged-for-$DISTRIBUTION >> version of vdr since there's no guarantee that the upgrade won't be >> scheduled for when nothing is being recorded. > you mean, a way to kill the VDR process only when there's no background > activity? Why not just send a power key via SVDRP? That doesn't guarantee that vdr will exit promptly or as soon as it becomes inactive. A different port may have been chosen, SVDRP may have been disabled, or the suspend patch may be in use (and my vdr package has this patch). There's "restart on becoming inactive" too, but I think that the restart part of that is best implemented in runvdr since it too may need to be restarted. > For the purpose of installing updates, I wouldn't like it if the update > happens when I'm watching live TV either, so we're at the same point again: > Shut down when no user is active and no background activity is going on. > Could be done by abusing the existing shutdown script. Perhaps. You could have a look at my packaging: http://zap.tartarus.org/~ds/debian/dists/unstable/main/source/vdr_1.4.4-1.ds.diff.gz > Another possibility would be to query the active/inactive state by SVDRP > commands, either by VDR itself, or by a plugin. I still prefer kill(1) and kill(2) - no faffing around with sockets :-) -- | Darren Salt | linux or ds at | nr. Ashington, | Toon | RISC OS, Linux | youmustbejoking,demon,co,uk | Northumberland | Army | Say NO to UK ID cards. http://www.no2id.net/ Fast ship? You mean you've never heard of the Millennium Falcon?
