6 ms·

I’ll wait to see if they have proper “sleep” mode such that they don’t keep draining batter while not in use, particularly when running Linux. This is something

I’ll wait to see if they have proper “sleep” mode such that they don’t keep draining batter while not in use, particularly when running Linux. This is something that baffles me about non Apple Laptops how they just refuse to catch up to such a great benefit.. (alongside the trackpad experience which I can may be live with).

f_devd 3y ago
It might be a bit of tinkering but you can enable suspend-then-hibernate with logind/systemd which gives you a similar setup to Macbooks but slightly slower to wake from deep sleep as it needs to go through post/grub again. With the bigger battery provide and new CPU I do hope it lasts longer when idle though.
- sowbug 3y ago
  Does that require a LUKS unlock step?
  - f_devd 3y ago
    I believe so, I remember an article on hibernation+luks being more difficult to set up
    - dathinab 3y ago
      it mainly lacks good UI/UX tools to do so and an agreement on what the right approach is, especially if combined with raid without raid it's not too hard, just bothersome and finding the right information can be annoying the main gotcha is that lockdown mode is incompatible with hibernation (no matter if it's encrypted or not) and some distros enable it by default when secure boot is used, which is most times the case with a full disk encryption setup
- hutzlibu 3y ago
  "It might be a bit of tinkering" Thats an understatement. I never made suspend-then-hibernate work reliable on linux and I tried it over the years on different systems. So I just had developed the habit of always entering hibernation, if I intend to close the lid for longer. On windows suspend-then-hibernate never failed so far. Drivers are just the crux with linux in too many cases. Not much you can do about it, except reverse engeneering drivers as a side project.
  - vetinari 3y ago
    Suspend and hibernate are calls into ACPI, so no amount of driver work is going to help, when the firmware is responsible.
    - hutzlibu 3y ago
      You can also reverse engeneer the firmware if you like, but as a matter of fact, it always worked reliable with windows laptops and not one single time with linux for me (open for suggestions of stable hw). And I am not a kernel dev, but I am pretty sure drivers are in general involved with waking up and sleeping as well. There are just lots of things, that can go wrong there. One small bug somewhere can result in blocking the hibernation => result is a drained battery.
  - f_devd 3y ago
    For the framework 13th gen it only required running a fairly recent kernel (>=6.1) for it to work, but that's definitely hindsight
  - dathinab 3y ago
    I use it too it's not hard to setup and works well (with very few exceptions and if you use a "recent" enough linux kernel, depending on your hardware). Once hibernation is setup just set systemd [Login]HandleLidSwitch=suspend-then-hibernate and maybe configure the delay with [Sleep]HibernateDelaySec=5min. Hibernation most times works nice too (even with Full Disk Encryption (FDE)) there is just one gotcha, lockdown mode is currently fundamentally incompatible with hibernation (even through it's docs seem to imply something else, they refer to a not yet existing feature). And some distros enable lockdown mode by default if you use a proper FDE setup. Leading to a lot of confusion about hibernation suposedly not working anymore in linux or being incompatible with FDE.
consp 3y ago
Many laptops these days only implement s2idle and no proper s3 support which is the main culprit. My lenovo only does a day (at best) in s2idle under linux (windows is worse), when I switched it to s3 it does about 4 days as expected for the amd cpu. Fortunately this has a switch in the bios (which says windows won't work afterwards which is cow manure). You get the added benefit that windows can't phone home when you close the lid.
- kitsunesoba 3y ago
  My X1 Nano Gen 1 has the same switch in EFI. Glad it exists but yeah it’s weird that Lenovo suggests that using it breaks Windows.
  - aksss 3y ago
    I think the idea is that Windows will occasionally not wake up properly, leaving the user to “choke it out” (power button for 5 sec to force a hard shut down, power button to power up). Sorry if you had any unsaved work. Personally, Hibernate works great, doesn’t require any battery life to sustain it, and always comes back with state intact, so it’s my go to.
    - kibwen 3y ago
      > Hibernate works great Seconded. All of these modern "smart sleep" solutions are uniformly worse than just giving me a got-danged hibernate option.
      - zbrozek 3y ago
        Hibernate is just slow, especially if you dual boot and have to wait through grub.
      - chrismorgan 3y ago
        > uniformly worse Extremely false: hibernate is way slower. Sleep happens within a second and awakes within a second. Hibernate requires writing all your memory to disk; if you have, say, 15GB of memory in use and a disk that can read and write at 1 GB/sec (far from the fastest, but also extremely far from the slowest), that’s fifteen seconds. For many (probably most) people’s realistic situations, hibernate will take over a minute. (Heh, my own laptop takes 4–8 minutes to hibernate if its CPU is slowed to 400 MHz to minimise power consumption, which incidentally happens automatically when it’s almost down to 0% battery power, shortly after which it triggers a suspend-then-hibernate, and the extra time that makes it take it brings it perilously close to running out of power before it finishes. Not certain why limiting the CPU has such a massive effect, but I’ve compared it when manually inducing similar circumstances by giving the CPU an impossibly low power envelope with `ryzenadj --stapm-limit 100 --fast-limit 100 --slow-limit 100`, which limits all cores to 400 MHz, and yeah, that makes hibernate slow.)
bobmaxup 3y ago
Every time I put my Apple laptop to sleep for an extended period, I come back to it completely dead. It cannot even boot for an about an hour while it charges to a minimum level.
- dathinab 3y ago
  a Intel Apple laptop? because they kinda don't care much more about the intel devices they had produced
  - bobmaxup 3y ago
    Yeah, unfortunate I suppose.
bodge5000 3y ago
Maybe its just me, but I've found Macbooks to be worse for this I have a Macbook Air, M1, which I don't use regularly but turn off after every use. When I go to turn it back on again, the battery's completely dead. I haven't had that from any non-apple laptop for a long time, and I sometimes go +6 months without turning them on The trackpad though I'm in agreement with. I never understood why people use an external trackpad with their imacs until I used it on the macbook
- stefandesu 3y ago
  I feel like my M1 Air could live on for weeks in sleep mode, as long as nothing's plugged in. Maybe you could try Sleep Air [1] to find the culprit? [1]: https://ohanaware.com/sleepaid/ https://ohanaware.com/sleepaid/

f_devd 3y ago

It might be a bit of tinkering but you can enable suspend-then-hibernate with logind/systemd which gives you a similar setup to Macbooks but slightly slower to wake from deep sleep as it needs to go through post/grub again. With the bigger battery provide and new CPU I do hope it lasts longer when idle though.

sowbug 3y ago
Does that require a LUKS unlock step?
- f_devd 3y ago
  I believe so, I remember an article on hibernation+luks being more difficult to set up
  - dathinab 3y ago
    it mainly lacks good UI/UX tools to do so and an agreement on what the right approach is, especially if combined with raid without raid it's not too hard, just bothersome and finding the right information can be annoying the main gotcha is that lockdown mode is incompatible with hibernation (no matter if it's encrypted or not) and some distros enable it by default when secure boot is used, which is most times the case with a full disk encryption setup
hutzlibu 3y ago
"It might be a bit of tinkering" Thats an understatement. I never made suspend-then-hibernate work reliable on linux and I tried it over the years on different systems. So I just had developed the habit of always entering hibernation, if I intend to close the lid for longer. On windows suspend-then-hibernate never failed so far. Drivers are just the crux with linux in too many cases. Not much you can do about it, except reverse engeneering drivers as a side project.
- vetinari 3y ago
  Suspend and hibernate are calls into ACPI, so no amount of driver work is going to help, when the firmware is responsible.
  - hutzlibu 3y ago
    You can also reverse engeneer the firmware if you like, but as a matter of fact, it always worked reliable with windows laptops and not one single time with linux for me (open for suggestions of stable hw). And I am not a kernel dev, but I am pretty sure drivers are in general involved with waking up and sleeping as well. There are just lots of things, that can go wrong there. One small bug somewhere can result in blocking the hibernation => result is a drained battery.
- f_devd 3y ago
  For the framework 13th gen it only required running a fairly recent kernel (>=6.1) for it to work, but that's definitely hindsight
- dathinab 3y ago
  I use it too it's not hard to setup and works well (with very few exceptions and if you use a "recent" enough linux kernel, depending on your hardware). Once hibernation is setup just set systemd [Login]HandleLidSwitch=suspend-then-hibernate and maybe configure the delay with [Sleep]HibernateDelaySec=5min. Hibernation most times works nice too (even with Full Disk Encryption (FDE)) there is just one gotcha, lockdown mode is currently fundamentally incompatible with hibernation (even through it's docs seem to imply something else, they refer to a not yet existing feature). And some distros enable lockdown mode by default if you use a proper FDE setup. Leading to a lot of confusion about hibernation suposedly not working anymore in linux or being incompatible with FDE.

consp 3y ago

Many laptops these days only implement s2idle and no proper s3 support which is the main culprit. My lenovo only does a day (at best) in s2idle under linux (windows is worse), when I switched it to s3 it does about 4 days as expected for the amd cpu. Fortunately this has a switch in the bios (which says windows won't work afterwards which is cow manure). You get the added benefit that windows can't phone home when you close the lid.

kitsunesoba 3y ago
My X1 Nano Gen 1 has the same switch in EFI. Glad it exists but yeah it’s weird that Lenovo suggests that using it breaks Windows.
- aksss 3y ago
  I think the idea is that Windows will occasionally not wake up properly, leaving the user to “choke it out” (power button for 5 sec to force a hard shut down, power button to power up). Sorry if you had any unsaved work. Personally, Hibernate works great, doesn’t require any battery life to sustain it, and always comes back with state intact, so it’s my go to.
  - kibwen 3y ago
    > Hibernate works great Seconded. All of these modern "smart sleep" solutions are uniformly worse than just giving me a got-danged hibernate option.
    - zbrozek 3y ago
      Hibernate is just slow, especially if you dual boot and have to wait through grub.
    - chrismorgan 3y ago
      > uniformly worse Extremely false: hibernate is way slower. Sleep happens within a second and awakes within a second. Hibernate requires writing all your memory to disk; if you have, say, 15GB of memory in use and a disk that can read and write at 1 GB/sec (far from the fastest, but also extremely far from the slowest), that’s fifteen seconds. For many (probably most) people’s realistic situations, hibernate will take over a minute. (Heh, my own laptop takes 4–8 minutes to hibernate if its CPU is slowed to 400 MHz to minimise power consumption, which incidentally happens automatically when it’s almost down to 0% battery power, shortly after which it triggers a suspend-then-hibernate, and the extra time that makes it take it brings it perilously close to running out of power before it finishes. Not certain why limiting the CPU has such a massive effect, but I’ve compared it when manually inducing similar circumstances by giving the CPU an impossibly low power envelope with `ryzenadj --stapm-limit 100 --fast-limit 100 --slow-limit 100`, which limits all cores to 400 MHz, and yeah, that makes hibernate slow.)

bobmaxup 3y ago

Every time I put my Apple laptop to sleep for an extended period, I come back to it completely dead. It cannot even boot for an about an hour while it charges to a minimum level.

dathinab 3y ago
a Intel Apple laptop? because they kinda don't care much more about the intel devices they had produced
- bobmaxup 3y ago
  Yeah, unfortunate I suppose.

bodge5000 3y ago

Maybe its just me, but I've found Macbooks to be worse for this I have a Macbook Air, M1, which I don't use regularly but turn off after every use. When I go to turn it back on again, the battery's completely dead. I haven't had that from any non-apple laptop for a long time, and I sometimes go +6 months without turning them on The trackpad though I'm in agreement with. I never understood why people use an external trackpad with their imacs until I used it on the macbook

stefandesu 3y ago
I feel like my M1 Air could live on for weeks in sleep mode, as long as nothing's plugged in. Maybe you could try Sleep Air [1] to find the culprit? [1]: https://ohanaware.com/sleepaid/ https://ohanaware.com/sleepaid/