It’s interesting this is still a problem going on twenty years.
I guess there’s some reason it’ll never actually work a linux laptop.
I guess there’s some reason it’ll never actually work a linux laptop.
Except it clearly can; every thinkpad I've owned handled S3 flawlessly in Linux, and Chromebooks manage fine.
I do not want to discuss that, I just want to show the value of anecdotal evidence postings: zero.
I’ve had thinkpads that don’t work either. The issue is suspend still doesn’t work reliably going on twenty years - even in the case where the laptop is explicitly a Linux laptop made for this purpose.
It means my default assumption has to remain that it probably won’t work (or at least I shouldn't expect that it will). I’m tempted by the framework because I like what they’re trying to do, but if suspend is still an issue it’s imo not worth it (even ignoring the bad battery tradeoff). That’s even ignoring the continued distance apple keeps putting between themselves and everyone else with their custom hardware.
Also the modular ports are essentially built in usb-c dongles. It’s not obvious to me why this is better than just having adapters. It takes up internal space and may be partly why the battery life is bad. Instead of adapters you have modules - is this a win?
On Windows sleep is a shitshow. It sometimes wakes up if I leave it alone during the night, it sometimes doesn't sleep at all (the screen stays on), it sometimes completely breaks (fans go full tilt, screen backlight comes on, but that's it) and I need to do a full reboot to make it work.
On Linux: press sleep, and it sleeps. Press power, it wakes up. If it sleeps, it stays asleep. When it turns on, the computer is actually usable, no forced reboot required.
This despite the fact that "HP recommends Windows 11", and the support page has no idea Linux even exists.
Still better than having to use Windows or OSX.
This is most likely wake timers. Some program or windows itself is waking your machine to do updates. Disabling wake timers will prevent anything from waking the machine while sleeping.
And this also happens during the day, say if I put it to sleep while going for a walk. It sometimes wakes up a few moments later, sometimes it'll sleep all the way.
My schedule is fairly regular, too, so unless the timers have large jitter, I'd expect them to trigger roughly at the same time (either after X time, or at Y hour).
This used to happen on Linux initially, too, and there were no timers there. I could stop it by disabling EC wake up. I've never found anything similar on Windows, so I just turn it off. At least it boots quickly...
powercfg /waketimers
powercfg /lastwake