4 ms·

The problem with that solution is most people want their PC to wake up when they wiggle the mouse.

LegitShady 4y ago
he didnt' say it shouldn't wake when it wiggles, he said they change the sensitivity required to wake it.
tpmx 4y ago
It turns out it's actually pretty easy to distinguish "wiggle the mouse" from "a truck just drove by nearby and triggered a one pixel movement". Microsoft should have quite a few employees that could figure this out.
- leni536 4y ago
  It is probably hard to do while the computer is asleep.
  - tpmx 4y ago
    As I wrote in another comment; it was possible to disable wake from sleep from within Windows for that particular mouse; that was promising. I don't know if it's possible to execute logic on mouse movements in the Windows sleep power state. Either way, I feel it's something MSFT should spend some time on, given the current situation.
    - sgtnoodle 4y ago
      If there is code running in that low power state, it's likely firmware on a microcontroller somewhere inside the motherboard chipset. Microsoft probably can't do anything to fix that, other than make more robust mouse wakeup a requirement for branding future products as "made for windows". Suspend isn't even usable at all on my desktop when running Linux. The AMD GPU doesn't come back up properly 3 out of 4 times, and then any userspace processes that were using any graphics APIs (such as the desktop environment) are borked. Fixing that would probably save me 1kWh per day. In my experience, I can't even get monitors to power off reliably, and I spend hours and hours fiddling with that sort of thing trying to debug it. Anything power-save related is a mess.
      - tpmx 4y ago
        If that's the case, at least they could make change the default of allowing wake-by-mouse-movement via a software update, as demonstrated by my experiment just now.
        williamscales 4y ago
        They could, but a lot of folks (anecdotally) do want their computer to wake on mouse movement. Maybe it could be tied to power management mode as a compromise.
        sgtnoodle 4y ago
        I remember a computer I had 15 years ago waking up on mouse movement, and it was great. My last couple desktops /might/ wake up on a mouse click, if they go into suspend at all on Linux. I assume the general regression in suspend in Linux is due to Windows mandating non-standards based power saving modes in the last decade.
More-nitors 4y ago
listening for clicking might be a better choice: 1. mouse doesn't have 'actively listen' for mouse movement - exceptions: ball-mouse 2. unintentional-clicking is much more rare
MereInterest 4y ago
There's a big difference between "wiggling the mouse", and having a non-zero reading from the mouse. With sufficient sensitivity, a breeze or vibration can cause the latter, but a wake-algorithm should be able to distinguish the two.
- withinboredom 4y ago
  Makes me wonder if there’s a huge power surge during late night earthquakes due to moving mice…

he didnt' say it shouldn't wake when it wiggles, he said they change the sensitivity required to wake it.

tpmx 4y ago

It turns out it's actually pretty easy to distinguish "wiggle the mouse" from "a truck just drove by nearby and triggered a one pixel movement". Microsoft should have quite a few employees that could figure this out.

leni536 4y ago
It is probably hard to do while the computer is asleep.
- tpmx 4y ago
  As I wrote in another comment; it was possible to disable wake from sleep from within Windows for that particular mouse; that was promising. I don't know if it's possible to execute logic on mouse movements in the Windows sleep power state. Either way, I feel it's something MSFT should spend some time on, given the current situation.
  - sgtnoodle 4y ago
    If there is code running in that low power state, it's likely firmware on a microcontroller somewhere inside the motherboard chipset. Microsoft probably can't do anything to fix that, other than make more robust mouse wakeup a requirement for branding future products as "made for windows". Suspend isn't even usable at all on my desktop when running Linux. The AMD GPU doesn't come back up properly 3 out of 4 times, and then any userspace processes that were using any graphics APIs (such as the desktop environment) are borked. Fixing that would probably save me 1kWh per day. In my experience, I can't even get monitors to power off reliably, and I spend hours and hours fiddling with that sort of thing trying to debug it. Anything power-save related is a mess.
    - tpmx 4y ago
      If that's the case, at least they could make change the default of allowing wake-by-mouse-movement via a software update, as demonstrated by my experiment just now.
      - williamscales 4y ago
        They could, but a lot of folks (anecdotally) do want their computer to wake on mouse movement. Maybe it could be tied to power management mode as a compromise.
        sgtnoodle 4y ago
        I remember a computer I had 15 years ago waking up on mouse movement, and it was great. My last couple desktops /might/ wake up on a mouse click, if they go into suspend at all on Linux. I assume the general regression in suspend in Linux is due to Windows mandating non-standards based power saving modes in the last decade.

More-nitors 4y ago

listening for clicking might be a better choice: 1. mouse doesn't have 'actively listen' for mouse movement - exceptions: ball-mouse 2. unintentional-clicking is much more rare

MereInterest 4y ago

There's a big difference between "wiggling the mouse", and having a non-zero reading from the mouse. With sufficient sensitivity, a breeze or vibration can cause the latter, but a wake-algorithm should be able to distinguish the two.

withinboredom 4y ago
Makes me wonder if there’s a huge power surge during late night earthquakes due to moving mice…