Here are the results of the little #Linux survey I ran early this week!
I must admit the results surprised me, I tried to explain them as best I could, but donβt hesitate to comment on the video if you have other ideas !
Also, the raw data is available from the link in the description :)
@thelinuxEXP I switched from i3 to sway many years ago. I had no problems with i3 but Xorg was a different matter. Wayland did scaling better: I configured it globally and it work for all clients (some blurry, most not). I used different monitors on my laptop in different places. Switching scales on Xorg was a pain. Wayland did per-display scaling, which I needed when using lodpi and hidpi displays together.
(IIRC, I switched in 2018)
@thelinuxEXP great Video, I like how you emphasize, that the results aren't representative and agree with most of your conclusions.
Just one thing: The number of answers are great, incredible. The amount is not the problem (scientifically), they aren't randomly chosen (as you mentioned yourself).
Still, I don't know a better approach as you did.
@thelinuxEXP I compared the stats with @gamingonlinux monthly survey (https://www.gamingonlinux.com/users/statistics/). Note that many (if not all) categories do not fully overlap and some entries are hidden in "other" categories. Still, it shows some differences between communities β and overlaps.
Nick's survey β GoL survey [%]
Distributions
Arch-based 33 β 39
Ubuntu-based 22 β 31
Fedora-based 21 vs 12
#Arch 22 β 21
#Fedora 19 β 11
#Ubuntu 9 β 10
#NixOS 7 β 1
#Mint 6 β 7
#Debian 5 β 5
#OpenSUSE 4 β 4
#Manjaro 3 β 9
Desktop environments
#Plasma 30 β 40
#Gnome 32 β 29
21% for "tiling WM" vs 9% for "WM only"
#Cinnamon 7 β 8
#XFCE 3 β 7
#MATE 1 β 2
#Cosmic 3 β 1
#Wayland 66 β 33
GPU vendors (assuming dGPU is dominant and reported in GoL)
#Nvidia 37 β 43
#AMD 39 β 53
#Intel 24 β 4
CPU vendors
AMD 50 β 67
Intel 49 β 33
Proprietary Nvidia drivers
87 β 98
Dual boot
nope 63 β 74
yes, #Windows 30 β 24
yes, #MacOS 1 β 0
other 6 β 2
I should have specified Ubuntu-based includes Ubuntu, same for Fedora-based and Arch-based. I forgot one more interesting point.
#Steam OS 0 β 2
@thelinuxEXP People shouldn't br surprised about Arch. It's a meme, but it's also a damn good distro.
I think there is this false narrative that you can either choose Arch for bleeding edge packages or you can choose another distro for stability. It's just not true. My Arch install is almost 3 years old, and I've never had any major issues.
Also Arch packages often LAG behind distros like Fedora, so the idea that it's some bleeding edge system is also false.
@thelinuxEXP I recently forked Paper, an open source flatpak notes app, and released it as Folio for which I created an AppImage and a Snap as well.
I can pretty much back up the lack of interest in Snaps as in the month since I released the Snap I've only had 32 downloads, where as the Flatpak is over 2500 with just an extra week or two of release π€·
No way to track the AppImage of course, but that's part of AppImage's problem from a developers perspective.
@thelinuxEXP According to the results, this aged like milk:
Nick @ The Linux Experiment
JCM
Hugo ι¨ζ
Keywan Tonekaboni
Uditt Lamba 
matthew
Ponda
Maxi 12x π
Fingel
Greg Ross
Q. Edwards
kiaanthelinuxguy
FineFindus
jstm ausm paralleluniversum
postnick
Lightπ§β
Mark Jones
First_Thunder
LogicalErzor
neopolitan
Patrick