Nobody should be surprised about the contents of this article. If there is anything I've learned from the years of being involved in the Linux ecosystem and reading Linux news, it's that Steven J. Vaughan-Nichols (and ZDNet as a whole) is not known for writing articles which are well researched, well thought-out, or delivered without bias. Most of Steven's pieces are just pro-Linux Mint and anti-everything else. I have nothing personal against the author to be clear, the simple reality is they have a bias and ZDNet is one of those news outlets that prefers quantity over quality and to maximize clickbait to drive clicks, so there's a home for the author to "spew" out articles.
Linus drives click. When it came to his view on the current situation on the desktop, he said: "I still wish we were better at having a standardized desktop that goes across all the distributions. There's been some progress there. I mean, this isn't a kernel issue, this is more of a personal annoyance how the fragmentation of the different vendors have I think held the desktop back a bit. But there has been some progress on that front too with Flatpak and things like that. I'm still optimistic, but it's been 25 years and it's going to be another few years at least."
He also stated in response to the fragmentation of application container formats (AppImage, Flatpak, Snap) that: "That may be what Chromebooks end up doing, is that maybe that will turn into a de-facto standard for desktop applications. When Chromebooks just start running Debian packages or something. We'll see."
Source:
I respect Linus' views and thoughts on what he believes to be holding back the desktop. Literally nobody else has worked on Linux as long as he has, so to say he knows a "thing or two" about the ecosystem would be putting it mildly.
But the ecosystem has room for everyone's voice and I don't agree entirely with Linus' remarks. Personally, I do not believe the development of multiple desktop experiences / environments is what is holding Linux back. Each project has purpose for being, whether you agree with it or not, and each try to provide some value-add for their users. Where I think the Linux community falls short is maximizing the interoperability of desktop applications and desktop environments through the use of FreeDesktop standards / standards. There's been some progress made, some occasional standards / specifications written that are widely use, but in the almost 20 years that FreeDesktop has been around we don't really have that many specifications to show for it.
Some examples, not the exhaustive list:
- Standardizing Autostarting of Applications, which hasn't seen any changes to the spec since its initial draft in 2006.
- Desktop Entry Specifications. While this is a great specification and allows for extensibility, it's fairly common to see GNOME vendor keys such as
X-GNOME-Bugzilla-Bugzilla, or docpath stuff such as X-DocPath. I don't see a reason why we couldn't improve the specification to have a standardized way of specifying the Issue Tracker URL with optional products and components, and standard local and remote documentation URL keys.
- MPRIS D-Bus Interface Specification. There's a lot to love about the specification but its player interface could be improved to more easily expose MPRIS metadata and Xesam properties.
Something that is also really frustrating is the lack of a modern specification for desktop notifications. We sorta have a half-assed standard which most desktops support, but that lacks text input / response actions (so say you get an instant message, you'd be able to reply to those within the notification should a desktop support that part of the specification), urgency levels aren't granular enough, and most categories are vendor classes rather than having an up-to-date list of categories.
On KDE's part, they lack support for what I'd see as a considerable amount of the existing specification, you can see here. But they support things like quick replies.
Why is that not part of an upstream specification? We should not have an expectation that software developers implement multiple forms of the specification, whether they're building an app or a library. By having splintered specifications, we're making it more difficult to support multiple desktop experiences / environments and forcing developers to make a choice on what platform(s) they support, or putting the bar so high that they don't bother at all.
I obviously don't agree that Debian packages and Chromebooks are the standard going forward, nor do I really see application container formats being problematic for software developers. The easy part is software delivery, the difficult part is developing the software itself in a manner which maximizes the likelihood of you providing an experience across all desktop environments and all Linux-based operating systems. It's why we see so many Electron-based applications. You can say what you want about "well it doesn't look native" or "it doesn't fully interop with my desktop" but it runs and likely operates in a consistent manner everywhere, which can't be said about GTK3 or Qt5 applications.
For software delivery, the reality is Snap has won. You don't see large swathes of developer marketing stories comes from the AppImage or Flatpak camps, nor the excitement from software vendors on those formats. But you do see them from Snap. In my opinion, the only reason Flatpak has had the success it has, is because it's developed by GNOME and Red Hat and force-fed to you through GNOME Software (something we fortunately do not have) and worse, GNOME Shell.
Back to the ZDNet article, let's touch on a few points:
None of the major Linux distributors -- Canonical, Red Hat, SUSE -- are really all that interested in supporting the Linux desktop.
I'm sorry but that is bullshit. Plain and simple. All of the above mentioned vendors invest huge amounts of money in desktop development efforts.
True, the broad strokes of the Linux desktop are painted primarily by Canonical and Red Hat, but the desktop is far from their top priority.
Red Hat heavily invests in Red Hat Enterprise Linux, GNOME and RHEL have massive deployments in enterprises, Point-of-Sale devices, etc.
Another major player in setting the tone of the Linux desktop are the smaller Linux communities. These include Linux Mint, Manjaro Linux, MX Linux, elementary OS, and Solus. They're all doing good work, but they're also running on a shoestring basis.
This might come to a surprise to Steven, but just because a project gets money does not mean it's going to put it entirely into the desktop. Some of us care about tooling and opt to invest there as well.
Take Mint, my own personal favorite desktop.
Case-in-point about constant Mint bias.
Steven goes on to talk about frankly non-issues like various Mint developers ensuring that their non-OSS work takes priority and various development items being deferred, such as Muffin work. None of it really fits into the picture of desktop fragmentation, but it plays into the narrative that Steven is trying to push that we're all in trouble.
For the million or so of us, there are hundreds of millions who want an easy-to-use desktop that's not Windows, doesn't require buying a Mac, and comes with broad software and hardware support. Are you listening Linux Foundation?
Such a desktop, in turn, would be more commercially successful than our current hodgepodge of desktops. This would mean that many more Linux desktop developers could make a living from their work. That would improve the Linux desktop overall quality. It's a virtuous cycle, which would help everyone.
In the case of the consumer options for devices they can go out, purchase, boot up and use, the issue isn't the lack of a singular "standard" desktop environment that is keeping people away. The entire issue around surfacing available devices for people to buy can be solved with marketing. We have had big names for years that ship computers with Linux or no operating system at all, whether it's Dell, Lenovo, even HP. Whether they actively market those devices to the general public is another story.
tldr;? No, the Linux desktop is not in trouble. Typical FUD non-sense. Move along.
