This article is more than 1 year old
Unity and Trinity: New releases for forks of abandoned Linux desktops
True believers really can bring about resurrections, it seems
Two minority Linux desktop environments were updated this week. Coincidentally named Unity and Trinity, both are forks that continue projects long ago abandoned by their creators.
With public testing announced, 7.6, which is the first update to Unity in six years, starts the process of modernizing Ubuntu's former desktop.
When it was still part of the Ubuntu distro, there were three separate branches of the Unity desktop. The original Unity desktop derived from Ubuntu's special launcher for netbooks, and was implemented as a plug-in for the Compiz compositing window manager – itself now discontinued.
In the meantime, though, he and collaborator Khurshid Alam are working on fixing some of the issues with the old Unity7 codebase.
Both Unity7 and the planned UnityX are entirely separate and independent from Unity8, Ubuntu's planned touch-driven mobile desktop.
Part of Ubuntu's problem was the size of its endeavors. Unity8 ran on the Mir display server.
This meant that along with both a desktop UI and separate a mobile UI, plus the Snap packaging format, Canonical was also working on a new display server to replace the now elderly X.org – duplicating much of the Wayland project.
There was once a separate non-composited Unity2D, like KDE built using the Qt toolkit. Unfortunately, it used Qt 4, unsupported since 2015, and the effort needed to move it to Qt 5 or the new Qt 6 would be very considerable.
Speaking of KDE, the other desktop seeing some love is a new maintenance release, version 14.0.12, of the Trinity project, a fork of KDE 3. KDE 4 appeared in 2008 and even then annoyed our own Steven J Vaughan-Nichols, and he was not alone.
As well as moving to newer versions of Qt, KDE 4 and later 5 brought in new facilities, such as "Plasmoid" desktop widgets. However, the newer versions also dropped basic functionality that some users liked, such as the ability for a single panel to span multiple monitors. Even if you never wanted that, there's an obvious comparison with Windows 11's inability to have a vertical taskbar.
KDE remains, as it has always been, highly customizable, but this shows an important lesson: one person's unimportant frippery is someone else's critical functionality. Freeing widgets from the panel and allowing them to run on the desktop may make for impressive tech demos, but it annoys users who just want to span their panel across the bottom of two screens.
With a widescreen display, horizontal space is plentiful, but vertical space becomes a premium feature. For the author, the way that Unity combined the top panel with the menu bar was a vital space-saver on a netbook and remains very useful today, when every monitor is a widescreen. If you have half a dozen windows, each with its own menu bar, that wastes a lot of vertical space, while Unity conserves that. In-window menu bars also require precision mouse control, whereas the ergonomic principle called Fitt's Law means that a screen edge is very easy to hit.
- Problems for the Linux kernel NTFS driver as author goes silent
- Pop!_OS 22.04: New kid on the Ubuntu block starting to show real muscle
- Fedora starts to simplify Linux graphics handling
- Fancy a remix? Ubuntu Unity and Ubuntu Cinnamon have also hit 22.04
GNOME eliminates this issue by simply removing menu bars altogether. However, concealing functionality behind hotkeys and hamburger menus leads to more time wasted hunting for functionality that used to be in plain view.
For the author, Unity combined the best of the Windows UI, such as window manipulation and a menu bar that could be controlled by standard keyboard shortcuts (alt+F for the File menu, for instance) together with macOS's efficient space usage.
The vertical app launcher used some of that plentiful screen width, while letting me see more of the document I was working on. Windows 11's mandatory horizontal taskbar, and the fat, graphical ribbons of modern Office apps, squander that precious space on a beginner-friendly UI, one that both gets in my way and leaves me less available workspace.
These may seem like small things, but such small paper cuts add up. They are the reason that whole teams of people get together and fork abandoned projects to keep them alive.
Today, there are half a dozen teams all maintaining separate Linux desktops that replicate the Windows 95 UI. KDE and Trinity, plus LXQt, do it using the Qt toolkit. Xfce replicated it using Gtk 2, while MATE forked GNOME 2; both are now moving to Gtk 3. Raspberry Pi's PIXEL forked and continues LXDE, while Mint's Cinnamon replicated the look and feel by forking GNOME 3.
They're all in danger of being left behind as the GNOME Project moves on with Gtk 4. And let us not forget that the reason that the GNOME project began was not for any functional reason: it was because KDE used Qt, which back then was not GPL-licensed software, and it was mostly written in C++, while Unix traditionalists tend to favor plain old C.
Elementary OS's Pantheon desktop is superficially macOS-like, but lacks functionality like a global menu bar – or menu bars at all – meaning that its top panel is almost entirely wasted, empty space. That helps to motivate Unity, but there are currently two separate Unity-derived codebases moving forward (both versions 7 and 8), and possibly a new UnityX to replace Unity 7.
That's a huge amount of duplicated effort. Meanwhile, Linux desktop market share remains tiny, though ChromeOS is selling strongly.
ChromeOS, like macOS, has none of this rich functionality or customizability. And yet, Macs far outsell all desktop Linux distros put together, and ChromeOS until recently sold more units than Apple Macs.
Surely there's some lesson to take away from this. What was that old saying? Something about divide and conquer? ®