Risc-V Mixer part 6

· Recent Quackings

Desktop Environment Selection and Configuration

Table of Contents

1#  ____  ___ ____   ____   __     __  __  __ _____  _______ ____
2# |  _ \|_ _/ ___| / ___|  \ \   / / |  \/  |_ _\ \/ / ____|  _ \
3# | |_) || |\___ \| |   ____\ \ / /  | |\/| || | \  /|  _| | |_) |
4# |  _ < | | ___) | |__|_____\ V /   | |  | || | /  \| |___|  _ <
5# |_| \_\___|____/ \____|     \_/    |_|  |_|___/_/\_\_____|_| \_\

Selecting a Desktop Environment for the Pinetab-V Tablet. #

So, now that the system is stabilized and up to date with all the modern applications we like to use, it is time to settle on a graphical desktop environment. Whichever environment we choose, there are some features needed specifically because the Pinetab-V is a tablet 2in1, and some features we will target out of our own preferences. These features will be listed below.

Mostly, because we are already at the end of this journey and know the answer, but many of these features are quite lofty, and one in particular is not a possibility at this time.

What we have tried so far. #

The device comes with what appears to be Gnome PaperWM, this was immediately removed simply because the "big name" desktop environments are absurdly bloated, and how Gnome cycled the workspaces along with the application menu was rather annoying. Gnome PaperWM also consumed a significant amount of system resources which at times rendered aspirations of productivity a fading dream.

What we wanted to replace it with was an environment tailored for 2in1 tablets, but we were willing to settle for an environment which was "highly modified" to handle the 2in1 features.

Paperd #

After learning about Paperd from an online article reviewing graphical environments for 2in1s we felt like it should be a good candidate for trying out. The review was encouraging and the interface was uniquely intuitive. Paperd was specifically written for 2in1 tablets, and to provide a productive graphical environment for 2in1 tablets that maximized the fullness of screen resolution for the user.

What we liked: #

What we didn’t like: #

Summary #

Paperd was well designed and regardless of it’s kinks, it was implemented successfully. Undertaking the design and build of a desktop environment is perhaps one of the most daunting feats to attempt. Our opinion is no discredit to the developers. It just appears towards the end of development, the developers were burned out, and chose to sacrifice project goals in favor of completing the project. Which is inevitably to their own credit.

Wayfire #

Since Wayfire was required in order to run Paperd, it was given a try to see if it could provide the desired features without the instability of Paperd.

What we Liked: #

What we didn’t like: #

Summary: #

In the end, the curve of having to relearn how to configure yet another desktop environment is what sealed the fate of wayfire for us. The experienced lag was did not help either.

Sway #

Sway was installed merely on a whim, and was intended to be used as a backup environment until a different environment could be settled on. I3 is our daily driver, and sway being nearly identical was a game changer, plus we used sway for years on a development box at the office. We switched to Hyprland at the office, which we have never regretted. When using Paperd and Wayfire were abandoned, Sway quickly became the favored candidate, and it still is what we primarily use on the Pinetab-V.

What we Liked: #

What we didn’t like: #

Summary: #

Performance was really quite excellent. Launching the foot terminal was nearly instantaneous. It just lacks the desired tablet features wanted.

Niri #

From the impression gathered from various online articles, Niri was a close second to Paperd, and after looking into it more deeply, it boasted more popularity. Unfortunately, Niri was not available in the Debian package repository during the trial period, which meant it had to be compiled from source. Thankfully, all of it’s required dependencies were already installed, so it just needed to be built.

Execution of the built binary for Niri failed repetitively, without an obvious indication of what the problem could be. Which put an end to our trial with Niri. We probably could have invested more into troubleshooting what was causing it to fail to run, but seeing how the quest to find the perfect desktop environment was eating away at our time, we ended it there.

Hyprland #

Another window manager we tried, and had previous postitive experience with was Hyprland. Hyprland is also currently being used on one of our development boxes, and we also maintain a configuration set for hyprland on github for.

This window manager failed to run with default settings as well, which was a disappointment, and meant it too would be relegated to the scrap heap.

Lomiri #

The Lomiri desktop promised to be almost exactly what we were looking for. It was claimed to be a lightweight convergent desktop environment with a growing community. It also was easily installable from packages in the debian repository.

What we Liked: #

What we didn’t like: #

Summary #

It would have been nice to have seen more of Lomiri than the login screen, a desktop wallpaper, and an unresponsive top bar. Unfortunately, what little we did see, Lomiri comprised of all the same bloatedness we have come to disdain about the populous desktop environment solutions, and since our exploration was limited by time, Lomiri was not a feasible choice.

Obvious Anticlimactic End #

It doesn’t matter how fancy a window manager is, if for whatever reason the system severely lags, the window manager is useless. Full stop.

Only one of the tested environments displayed reasonable performance, and that was Sway. So, without any better alternatives, we were swaying again. (Terrible joke...)

#riscv #tablet #desktop environment #window manager #environment #customizations
last updated:

Home