On 28.01.14 at 13:48, Patrick Burroughs (Celti) wrote: > On Tue, Jan 28, 2014 at 8:51 AM, Daniel Landau <daniel.lan...@iki.fi> wrote: > >> On Mon, Jan 27, 2014 at 10:49 PM, Patrick Burroughs (Celti) > >> <celticmad...@gmail.com> wrote: > >> And the udev rule works! Now I can completely do away with > >> laptop-mode-tools. > >> > >> For the curious, I used: > >> SUBSYSTEM=="power_supply", ATTR{status}=="Discharging", > >> ATTR{capacity}=="5", RUN+="/usr/bin/systemctl hybrid-sleep" > >> > >> This should work for any laptop that emits battery uevents at a > >> reasonable rate; mine didn't seem to at first, but started sending one > >> per 1% drop after it dropped below 15% or so, so if you want to test > >> it you may be waiting a while. > > > > That's cool! Do you think this would fit somewhere in the Wiki or as an AUR > > package? > > A wiki article seems best, so that people can adjust the type of sleep > and percentage chance to best suit their laptop. >
It's already there: https://wiki.archlinux.org/index.php/Laptop#Udev_events You should also be aware of this caveat: https://wiki.archlinux.org/index.php/Talk:Laptop#RUN.2B.3D.22.3Ccommand.3E.22_won.27t_work_reliably > > Are you sure you want hybrid-sleep instead of plain hibernate: both > > will save your state to disk but with hybrid-sleep you run the risk of > > running your battery all the way down, which supposedly isn't healthy for > > the battery > > (http://www.pcworld.com/article/191574/long_live_your_laptop_battery.html). > > There is a small danger of that, but my laptop uses little enough > power when suspended that that 5% charge will give me hours to plug it > back in. The hybrid sleep is just a final failsafe and the hibernate > part should never actually be needed. > > Regards, > ~Celti
pgp8SOshwoLMq.pgp
Description: PGP signature