I've recently got a FreeSync monitor and am still figuring out how to get VRR (variable refresh rate, or "adaptive sync") to work consistently.
I'd love to hear your experience with VRR.
Some of my tests:
Sway
In sway
, I set adaptive_sync on
for my display, and swaymsg -t get_outputs
reports Adaptive sync: enabled
- Unfortunately, by default stuttering in games is present
- To fix, I had to use
sway -Dnoscanout
. Apparently, this has its drawbacks. - Even with
-Dnoscanout
, moving the mouse breaks VRR and causes stuttering
Hyprland
In hyprland
, I set vrr = 1
and get similar results as sway -Dnoscanout
- Fluent motion by default, but moving the mouse introduces stuttering
Gamescope
In gamescope
(embedded from VTT, with --adaptive-sync
), I get the best results yet
- Stable fluent motion
- Mouse doesn't break it
My setup is 6600xt, Gigabyte M28U monitor, Arch 6.1.64-1-lts. I test with vrrtest and with Ghostrunner on wine-tkg-staging-fsync-git 8.13.r7.gc210ef9f-327
Yes! I'm on an eternal quest to find games that will just shut up and let me play.
My most recent find is Ghostrunner. Starts with the CPU doing the first kill and then you're off.
Before that it was Celeste. I now realize in both games the player dies a lot. Maybe there's a correlation between how much fun I have and how much the game allows me to die without repercussions 😅
I guess I'll need to try Elden Ring now. There's gotta be dozens of us. Anyone have more recommendations?
For me it reached a point where I now expect a new game I'm trying to just work. This was a monumental shift when I first realized that a few months ago.
Your best bet is Steam/Proton, since Valve stands behind it and development on all the Proton components (Wine, DXVK, VKD3D, Gamescope, ...) is very active.
If you get games outside of Steam (I often prefer GoG if that's an option, plus I have some itch.io bundles purchased a while ago), some tinkering may be necessary. For those, I like to go "vanilla" with Wine(-GE-custom usually), plus DXVK or VKD3D on top. There's also Lutris to help with these scenarios. Works great too.
Another topic is native Linux games. There are some gems which work beautifully. I recently finished native Celeste from itch.io and it was flawless. Another great Linux port is Bastion. But some older titles may have compatibility issues - missing or incompatible libraries, broken gamepad support or stuff like that. For those, the Windows version via Proton may actually work better than the native version. Luckily, we can now pick either one.
This is excellent.
I'm happy to use Gamescope for many use cases but I've always had issues FSR upscaling Civ4 with it. The mouse cursor would misbehave. Tested just now with Wine-GE-Proton8-11 and it beautifully upscales from internal 1280x720 to my monitor's 2560x1440.
Can be turned off with WINE_FULLSCREEN_FSR=0
which seems to fall back to nearest neighbor scaling. That's fine for me.
I'm not sure if WINE_FULLSCREEN_FSR_STRENGTH
has any effect though. Maybe I'm just too blind to see the difference.
I'm very happy to see these developments in the Linux gaming world. Thank you to everyone involved!
Generally these UEFI logos are some kind of a standard - search for "UEFI BGRT" to find out more. But it's up to the vendor whether to allow (or usually not) changing them.
ThinkPads are somewhat special in this regard - the firmware package includes a README.TXT
which says:
* THIS VERSION OF THE FLASH UPDATE PROGRAM GIVES THE OPTION OF *
* REPLACING (OR ELIMINATING) THE DEFAULT "LENOVO" IMAGE THAT IS *
* DISPLAYED DURING SYSTEM START UP. *
...and then proceeds to list instructions which don't work 😉
I don't think Dell has this as a feature of the firmware update tool.
![the background blur](https://lemmy.ml/pictrs/image/30c4fb5d-a0a0-4f74-9a61-04dd53214522.jpeg?thumbnail=256&format=webp)
![](https://lemmy.ml/pictrs/image/30c4fb5d-a0a0-4f74-9a61-04dd53214522.jpeg?thumbnail=1024&format=webp)
I spent way too long to get it working and I was almost ready to give up.
I tried the standalone ISO updater with a LOGO.GIF
file. It recognized the image, but after reboot the original Lenovo logo was still there. No error message, no warning, nothing.
Same with the Windows firmware updater.
The one thing that finally worked was a Winflash64.exe
incantation found buried somewhere deep in the Lenovo forums 😠
the incantation
Winflash64.exe -patch -logo logo.gif -exit
Posting for information only. This worked for me, but may - and probably will - brick your machine. You have been warned.
![the background blur](https://lemmy.ml/pictrs/image/230c16aa-2901-4642-a5d8-c42d1f813b4d.jpeg?thumbnail=256&format=webp)
![](https://lemmy.ml/pictrs/image/230c16aa-2901-4642-a5d8-c42d1f813b4d.jpeg?thumbnail=1024&format=webp)
I've been trying to test Intel Rapid Start Technology on my X1C9 and see how it compares to os-driven hibernation.
But my firmware doesn't show the option for IRST. I've seen screenshots from other people having this in the Power section.
Is this even a thing on recent ThinkPads? Has anyone been able to enable it?
I have the partition with correct GUID type. I also have OPAL enabled on the drive, but I don't think it should affect this feature... or does it?
Interesting, hadn't heard of powerstat
. I'll be checking that, thanks!
Agreed, that sounds like the way to go. I was hoping there was already something to do the monitoring for me :)
Thanks, I do like powertop
. I think it's pretty good for short measurements, e.g. over 30 seconds:
% sudo powertop --time=30
The battery reports a discharge rate of 4.17 W
The energy consumed was 125 J
The estimated remaining time is 11 hours, 4 minutes
But in the real world I will not be getting 11 hours of runtime. The moment I start a browser or play a video, power consumption goes way up.
tl;dr: I'm looking for something like AccuBattery, but for Linux
What do you use to measure laptop battery life in Linux?
I can easily get a momentary estimate of battery life. But this fluctuates based on load, screen backlight etc.
I'm looking for something that will collect my usage patterns over time - load, #processes, screen backlight, ... - and allow me to predict remaining runtime more accurately.
I'd love for the data to be parsable, so that I can analyze it myself and e.g. find the "worst offenders" - processes affecting battery runtime the most.
Thank you for any tips!