I've had hibernation working well on a stripped-down/modified SuSE 9.3 based system for quite a while now but I had an idea on the way into work today and thought I'd try to get some suggestions.
The problem is that the M1-ATX (don't know about the Opus but I'm kicking myself a bit for not getting it instead basically for reliability issues) signals the motherboard power switch connector when it loses 12V on the ACC line like it's supposed to but will shut down hard if it receives 12V on the ACC line again before it finishes shutting down, so you can't just implement this in software or script modifications.
What I basically want to do is have the computer stay on for say 10 minutes and then hibernate if the car hasn't been turned back on or started within that time. It would just be nice for quick stops so the startup time would basically be zero instead of the 30 seconds or so it takes to resume from hibernation. Any thoughts?
This is rather interesting, I am using the M1-ATX myself and have been pretty happy with it. The few things I dont like are the limited range of automated power off options (45 seconds or never?!?!?). In theory the power off is nothing more than a command that is interpreted by the apmd to call shutdown. I assume that if you found where the 'hook" is you could implement your own script that would call hibernate instead of shutdown. This could be very handy. If I wasn't stuck in Windoz right now I would look into it.. If you dont have any luck with it I will check it out when I get back home and am not tied to this hotel's internet connection ;).
You're right about limited shutdown timing options. My problem though isn't hibernating as that works just fine and the "power button" signal gets sent properly and is interpreted properly. What I want to do is just delay hibernation for 10 minutes IF 12V doesn't show up again on the ACC line of the M1-ATX. The problem arises that if you do delay hibernation that means the computer is still on, and when you go to start the car again (applying 12V to the M1-ATX ACC line) it will shut down hard and then reapply power about 3 seconds later thus killing any delay you might have implemented in the shutdown/hibernation scripts.
I suppose I could put the M1-ATX in dumb mode and build my own shutdown circuit but it would be nice if I didn't have to.