Since time immemorial—or, more likely, the late 1990s—the intractable problem of “fragmentation of the Linux desktop” has been debated on the internet. While some contend that the wide variety of competing distributions offers more choice to users, that choice can also be overwhelming—making it too difficult for new users to decide on a distribution, or leading them to choose a distribution that is poorly-built or unsupported, providing a bad first experience.
While these arguments have merit, they ignore a critical problem: The infrastructure and developer attention needed to maintain a distribution is extensive, and difficult to justify. Long-running Linux distributions have stopped operations due to a lack of resources, and it is time for Linux Mint to consider doing the same in order to prevent developer burnout, while transitioning Cinnamon into being a fully distribution-agnostic desktop environment.
Popular Linux distributions are ceasing operations
Shortly after the release of Red Hat Enterprise Linux 8 in May, Fermilab announced that there would be no new edition of Scientific Linux, ending Fermilab’s 20-plus year history of maintaining their own Linux distribution. Scientific Linux is little more than a recompiled version of RHEL sources, with Red Hat’s trademarks removed. This strategy made sense at the time, as RHEL is a paid, commercial distribution. Red Hat’s 2014 acquisition of CentOS—a general-purpose free recompile of RHEL sources—made Scientific Linux functionally redundant, particularly with the introduction of CentOS Special Interest Groups (SIGs).
SEE: 10 ways to prevent developer burnout (free PDF) (TechRepublic)
CERN withdrew from Scientific Linux in 2015, beginning a migration to CentOS, with Fermilab announcing their own migration to CentOS 8 as part of the transition from their own distribution.
Likewise, the Arch-based Antergos distribution announced plans to shut down, as the developers “no longer have enough free time to properly maintain Antergos,” and that “continuing to neglect the project would be a huge disservice to the community.” Certainly, community members have already announced their intent to continue under the name Endeavour, which will be a significant undertaking—building a user-friendly installer for Arch is an interesting science experiment, considering that this is more or less at odds with Arch’s goal of reducing abstractions that complicate system management.
Why Linux Mint became popular
Linux Mint has the unique distinction of being pragmatically correct twice, relative to the history of Linux on the desktop. When Mint was introduced in 2006, patent-encumbered codecs were not straightforward to install in popular distributions like Ubuntu or Fedora; likewise, proprietary software like Adobe Flash required separate installation, which was itself often a challenge. In part, this was made possible due to Mint being distributed from the EU, where software patents are essentially unenforceable.
Circumstances changed shortly thereafter, as Ubuntu added an extra screen to the installer to install codecs starting with Ubuntu Linux 7.04, and in 2008, various third-party repositories for Fedora merged to form RPM Fusion, providing a single source for packages not provided by Fedora for legal reasons. By 2010, Google Chrome 5 was released, providing an embedded Flash plugin, and support for Linux (and Mac OS), making the process of using Flash on Linux more straightforward.
Since then, Flash adoption has plummeted, with support ending at the end of 2020. Patents for MPEG-2, MP3, and Dolby AC3 have since expired, allowing Linux distributions to provide this capability freely, out of the box. While Mint was the first distribution to effectively solve this problem, the conceit that Mint is easier to use because it provides codecs installed by default no longer holds merit, as other distributions have since caught up. Mint has actually regressed in this position, as codecs are no longer installed by default as of Linux Mint 18, making this identical to other Linux distributions.
The desktop environment debacle of the early 2010s
Nearly simultaneously, every major OS made highly polarizing changes to the user interface. Microsoft introduced the “don’t call it Metro” interface with Windows 8, in 2012, which landed with a thud, and in part, prompted the exit of Stephen Sinofsky. In 2014, OS X Yosemitie attempted to make Helvetica Neue the default font, and abandoned the idea a year later.
SEE: How to manage job stress: An IT leader’s guide (free PDF) (TechRepublic)
Linux had their own schism, for desktop environments. Ubuntu Unity, originally developed for Netbooks, was introduced in 2010 to widespread derision, though had redeemed itself by the release of 12.04, with TechRepublic’s Jack Wallen migrating back to Ubuntu after jumping ship following issues with the initial releases.
Likewise, in 2011, the introduction of GNOME 3.0 on Fedora 15 was met with derision. GNOME 3 was intended for use with touchscreens, upending the usage patterns that users had become familiar with, prompting Linus Torvalds to declare it “unacceptable.” Dirk Hohndel, then-chief Linux and open-source technologist at Intel, declared at the time that “Gnome 3 is just completely unusable as far as I’m concerned.”
None of these were nearly ready for primetime when they launched, and this drove users away. For a time, Linux Mint “just worked” in a way that other distributions struggled to do, because they pushed too-new software on users. Out of this chaos was born Cinnamon, the fork of GNOME 3 built for Linux Mint that uses the classic desktop paradigm introduced in Windows 95. It’s familiar, and that’s a good thing.
Cinnamon, the raison d’être of Linux Mint
Cinnamon’s familiarity to millions, and the easy learning curve it provides by retaining a usage paradigm nearly 25 years old, is necessary, in a way that proponents of GNOME or KDE may be unwilling to admit. While Cinnamon is not the only desktop environment shipped by Mint, the distribution has jettisoned the KDE edition with the release of Mint 19. While Mint did not start with Cinnamon, for some time the bulk of original code produced by the Linux Mint team relates to Cinnamon—it is the reason the distribution has enduring popularity.
That said, the process of developing a Linux distribution and developing a desktop environment are rather dissimilar. Clément Lefèbvre, the founder and project leader of Linux Mint, does a fantastic job of guiding development of Cinnamon, though noted his own frustrations with the project in the March Mint update. The post is difficult to summarize succinctly, though he notes that “I personally haven’t enjoyed this development cycle so far,” and notes a divide between the concept of “users” and “developers.”
The following month, Lefèbvre—who simply goes by Clem, in the Linux community—walked back the comments noting that he is not “depressed,” despite some blogs reporting it as such, adding that “I also talked a tiny bit too much about what was going on within the team. On the one hand it is part of my role to report on the progress being done, on the other hand we’re dealing with individuals, there are people involved, efforts being made, feelings which can be hurt and it’s part of my role also to protect that.”
Clem doesn’t need to carry the world on his shoulders
Linux Mint is actually two distributions—the Ubuntu derivative, for which Cinnamon, MATE, and Xfce editions are provided, and the Cinnamon-based Linux Mint Debian Edition (LMDE), which exists “for the Linux Mint team to see how viable our distribution would be and how much work would be necessary if Ubuntu was ever to disappear.” Notably, LMDE previously had Xfce and MATE editions, though those have been jettisoned as part of an increased focused on Cinnamon.
Ubuntu is practically in the too-big-to-fail category, as Linux distributions go. While Canonical has abandoned development of Unity for Ubuntu—switching back to a modified GNOME 3—the distribution is continuing. Canonical is, at a minimum, solvent—particularly as expenditures for development of Unity stopped as programmers on that project were largely laid off.
SEE: Startup Republic: How France reinvented itself for the 21st century by wooing entrepreneurs to Paris (cover story PDF) (TechRepublic)
Ubuntu is not going anywhere. But, that only addresses why LMDE is unnecessary, not Mint overall. Maintaining this parallel plumbing for an alternative Mint for a doomsday scenario is paranoia, but it surfaces an interesting point: Cinnamon is, to an extent, developed to be distribution-agnostic, partially as a consequence of the existence of LMDE. Most of the original development for Mint is focused on Cinnamon, though maintaining the plumbing for the Ubuntu and Debian-based distributions—and other infrastructure, such as the website—is a massive undertaking, and a time sink for a team this small.
Cinnamon has momentum behind it, as the progressive, feature-rich implementation of the classic desktop paradigm for Linux users. (For comparison, MATE—while venerable—is essentially in maintenance mode.) Persisting in maintaining Linux Mint as a platform to showcase Cinnamon makes no sense, when the labor of maintaining a distribution is handled—better—by Ubuntu, Fedora, SuSE, and Arch, among a select few others.
Ultimately, the benefit of Cinnamon can be realized as a truly distribution-agnostic desktop environment. Most of the work is already done: Fedora already has a Cinnamon spin, and can be installed in Debian, OpenSuSE, and Arch (among others). Transitioning Linux Mint development efforts to make Cinnamon an Ubuntu Flavor—adhering more tightly to Ubuntu’s infrastructure and release timelines, rather than operating independently and running the risk causing package conflicts—would deduplicate a great deal of work, providing more time to further improve Cinnamon, and ease the strained schedules of Clem and other Linux Mint contributors.
For more on Linux, check out “Fedora 30 brings immense quality of life improvements to Linux on the desktop” and “Half of employees think the cloud is actually in the sky, according to a third of IT workers” on TechRepublic.