I guess this is more of a question of where do I go to find out when xrandr version 1.5.1 will be published in Ubuntu? It’s already available in Arch Linux and was released in August 2019. There is a bug from 2010 I want to have fixed.

Ubuntu 16.04.6 LTS current version is:

$ xrandr --version
xrandr program version       1.5.0
Server reports RandR version 1.5

I’m not well-versed on the subject but could I simply get 1.5.1 source code and compile it? Or should such a mission critical app such as xrandr never be compiled from source?

TL;DR Why it matters

Everyone can try these short little tests on their platform to see the importance of xrandr version and the gamma bug.

The current Ubuntu version has the bug that’s been around for 9 years:

$ xrandr --version
xrandr program version       1.5.0
Server reports RandR version 1.5

Basic problem is xrandr reports the wrong gamma values:

$ xrandr --verbose | grep ^DP-1-1 -A5
DP-1-1 connected 3840x2160+1920+0 (0xa5) normal (normal left inverted right x axis y axis) 1600mm x 900mm
    Identifier: 0x43
    Timestamp:  538179391
    Subpixel:   unknown
    Gamma:      1.0:1.1:1.3
    Brightness: 0.63

My “redshift-like” application has set gamma to Red = 1.0, Green = .88 and Blue = .77 but RGB is incorrectly reported as 1.0:1.1:1.3. Now imaging we want to increase brightness to .65. If we don’t change gamma at the same time existing settings for gamma are reset to 1:1:1. So we pass what we think are the current values:

$ xrandr --output DP-1-1 --brightness .65 --gamma 1.0:1.1:1.3

Low and behold the screen goes super bluish-greenish and kills our nighttime settings for reddish hue. When we check current settings again we find the values are inverted again:

$ xrandr --verbose | grep ^DP-1-1 -A5
DP-1-1 connected 3840x2160+1920+0 (0xa5) normal (normal left inverted right x axis y axis) 1600mm x 900mm
    Identifier: 0x43
    Timestamp:  541629314
    Subpixel:   unknown
    Gamma:      1.0:0.91:0.77
    Brightness: 0.65

So no matter what value xrandr --verbose is reporting we always have to use 1 / gamma to get real gamma on Red, Green and Blue channels. After correcting our code, we have to put in a test for version 1.5.1 to not correct our code and use the gamma values returned. Assuming the bug has been fixed in version 1.5.1 which I have yet to compile and test.