Updated: April 8, 2017
Several days ago, Mark Shuttleworth publicly announced that Canonical will stop the development of Unity8, the phone and the whole convergence idea. This seemingly sudden and possibly shocking change will come into effect in 2018, with the next LTS release.
By now, you've heard and read a lot of rumors and stories, analyzing this new situation, the future of Ubuntu as an operating system, and what all this means for us, Linux folks. Well, rather than quoting snippets from the Ubuntu Insights news article, I will focus on what the technical and strategic aspects of the change mean, and why you should be worried.
There's a lot of history and emotion behind Unity as a desktop choice. If you remember, when I tested UNR 10.10 Maverick, I had my first taste of what Unity was going to be. I wasn't happy, and I foresaw many people rebelling against this new pseudo-touch framework. At that time, Gnome 2 was still around, simple and sweet and functional to the max, and little did I know that its successor, also pseudo-touch, was going to suck balls, as I've outlined in my first Gnome 3 review. Ah, the memories.
Since, Unity became the flagship desktop, and Ubuntu grew, offering us decent releases with a semblance of professionalism that the rest of the Linux world lacked. But then, like any project too big for its shoes, it stalled and stumbled and lost momentum, not helped by the likes of systemd or Wayland, and the infinite fragmentation in the Linux world in general. Why, well, we shall get to that shortly.
In parallel, Canonical tries to develop its mobile platform. The first noble attempt was Ubuntu Edge, and I was super excited about it, I even ran a gadget competition of my own, alas this one never came to be. Then, we finally got the Aquaris phone, and eventually, the nice M10 tablet, which came with a rather sleek, smart yet ultimately incomplete idea of mobile-desktop wedlock, titled convergence. Big dreams, great ambitions, but they need solid foundations.
Why Unity failed
As Ubuntu grew and became more popular so did resistance to the attempt to make the hobbyist concept that was the Linux desktop into something commercially viable. Canonical gave us online searches, but people clamored their privacy was at stake. We had the Ubuntu Store and you could buy software through Ubuntu Software Center, but then this one disappeared, too, replaced with a semi-abstract Teletubby-happy nonsense called Gnome Software Center.
Ubuntu 14.04 was and still is the finest Ubuntu release, followed by a series of mediocre offerings. Xerus was a very weak, uninspired LTS successor, and when you combine all the different elements, it becomes apparent that the Canonical ship lost wind roughly in 2014.
Mir didn't take off as quickly as it should have, Wayland had its own challenges, people were not buying into the Web apps, because they were slow and cumbersome compared to native applications used by iOS and Android, and of course, without a financial case, no hardware vendor was going to invest time supporting Ubuntu on their mobile devices. That also means no suitable drivers or software, and without strong apps, touch platforms are pointless. Even Microsoft is struggling in this space, for this very simple reason, despite huge resources and intellectual capital.
Not helping was the entire community, which preferred to do things not-the-Canonical way. Linux Mint is a direct response to this dislike, as well as half a dozen other distros. The additional fragmentation further eroded the use case for Unity. Plus, of course, the introduction of new, complex and rather pointless frameworks made the entire Linux world falter, and we've seen a significant deterioration in quality and enthusiasm across the board. Hardly the loving, creation-fostering environment for greatness.
But the resistance is an excuse, not the reason. Canonical did not have the necessary breadth and depth of innovation to compete against the giants, because Linux is a terrible development platform for commercial use. For every nice thing that Ubuntu had, Android had nine. And even so, despite its enormous commercial success, Android is not good enough for Google. It's developing Fuchsia, a new operating system, because like Apple, it understands and knows you MUST own the entire stack if you want to be successful, as I've outlined in my professional distro recipe. It's the only way.
And what do you have in the Linux desktop world? Fragmentation. Sadness. A million apps being redesigned for the sake of it, completed to 93% of the original version and then abandoned when their devs get bored. Libraries on Ubuntu do not run on Fedora and vice versa. Three hundred different everything for everything, with zero compatibility.
Wait. Before you say - Red Hat or SUSE - the enterprise distros offer 10 years of support, they promise ABI compatibility (unlike Debian, which isn't used in enterprises), and they narrow down their support to a select list of software, like various compilers or CAD tools. Even certain filesystems aren't fully supported. For instance, Ext4, which home users love and whatnot, is still considered experimental in the big world. In the best case, you get read-only support. Because serious companies don't play stupid, costly games.
Back to the desktop. Ubuntu on mobile meant development frameworks that would make it simple and easy to port, test, deploy, etc. But these do not really exist. This is why we see the likes of Snap becoming more and more of a reality, however, that does not help Canonical right now. If you want to read more on where Unity8 is, please do. Just an alpha really.
Faced with inability to compete with Android, Ubuntu everywhere just could not happen. When your commercial software success relies on immature IDE, amateurs and volunteers doing stuff in their own time, then it just isn't meant to be.
The way forward - Gnome
Smartly, Canonical is stopping a pointless Utopia project and going back to what it once did so well. The Desktop. Cutting on the development of all these other ideas should allow Ubuntu to become a better product. And maybe even a commercially viable product.
This needs to be happen fast - because once Google lands on the desktop, and not just with Chrome OS (and Chromebooks, so to speak), but something much more serious, there won't be any chance left in this area, either.
The choice of the alternative, though, is dubious. Gnome 3.
Is this a good thing?
No, it is not. If you look at Gnome 3 as a commercial product, it's a joke. Even now, approximately six years since its inception, it still does not offer minimize or maximize buttons by default, you can't really use your desktop except as a glorified wallpaper, and there's no easy access to application shortcuts or icons. You must use the extra mouse click to open the Activities menu and then launch your Favorites. This is 2x more actions than Gnome 2, Windows whatever or any other sane desktop requires. It is a conceptual regression that only works because there's no real company behind Gnome 3.
Now, recently, I've been having a lot of fun with Gnome 3. Fedora 24 and Fedora 25 are great examples. But that's just silly games. That's me testing and pushing the limits of the operating system, tweaking, playing, enjoying myself. A hobby.
If you think Windows 8.1 is stupid, then you should use Gnome 3 for a while. This sounds like an insult, and maybe it is, but that's the simple, sad reality. You cannot fight against human nature, and any product that tries to complicate things does not deserve its place on the survival ladder. In 2017, Gnome won't let you right click and create a file by default. I'm not joking.
Or read my Gnome 3 accessibility guide to see how many little changes need to be added just to make the desktop sane. And we're not done, because I have another 2-3 articles in my publication queue, revolving around this very same topic.
So perhaps Canonical will be able to transform Gnome 3 into a usable entity. Yes, it can be done. But that will also be a compromise. Perhaps less stringent than having to rely on the goodwill of the entire community to understand your dream and follow suit, but it will still mean regressions, pointless choices and whatnot. Even Ubuntu today isn't immune from Gnome silliness. Just use the file manager for a few minutes, and you will see why.
The upcoming death of Unity also shows that the traditional Linux community is just not capable of adapting and evolving. In the current format, any attempt to make Linux appealing to the masses will fail. For ideological as well as technological reasons. The latter can somewhat be fixed. But the former, the only way around is to replace the audience and develop Linux for other people, who want consumer products rather than sandboxes. In the end, Canonical has to go for the lesser of evils, but an evil nonetheless.
But there's something good after all
Less fragmentation. Yes. Linux is all too heavily fragmented. 90% of it is simply unnecessary. As cruel as it sounds, I'd ax 90% of all applications, desktop environments and distros. Maybe then we could see high-quality products finally happen. Rather than have 100 three-man projects, we could have three 100-man projects. So much better and sustainable in the long run. Also, make sure no developer ever makes UI or product choices, but that's a fight for another time.
There's still time. Unity will remain supported until at least 2021, when the Xerus LTS expires. Four years is a very long time in the software world, especially Linux. Who knows what will happen till then. So we won't see Unity go away just there. It will be a gradual phase out. What we surely won't see is any new development on the phone, the tablet and the whole mobile space. I'm contemplating what to do with the devices I have.
So what will happen
OK, the last piece of this rant. The future. Where will Ubuntu be in five years? Same spot. Very simple. It will still be the most advanced desktop distro, with all the limitations of the Linux world. But without fundamental changes, it won't be able to rise above the crowd. It will be slowed down by half-dead projects with no patches, developers who will insist you change your usage rather than they change their code, regressions for no good reason, and so forth. And it will be running a desktop environment that won't even let you minimize your windows. Hey.
Maybe, just maybe, Canonical will be able to find innovative ways of adding the commercial spin to Linux, with extra focus and resources it has just freed from its mobile adventure. After all, it just needs Microsoft Office and Adobe Photoshop to succeed. Very simple.
But the way I see it, steady-state systems don't change. Unless you invest energy, nothing happens. With Linux desktop the way it is, still struggling to give you printing to Samba devices or codecs or whatever, and it's been like this or getting worse ever since I started testing distros (roughly 12-13 years ago), why should I believe there's going to be anything new in this space?
Perhaps the desktop will change - but it won't come from current distros. It's going to be Google, or a crazy miracle from Canonical. Looking at the distro landscape and the Ubuntu story for the past decade, I'm not fully convinced it will be Ubuntu.
Think what you will, but the ONLY player in the Linux desktop space that really tried to innovate was Canonical. Probably the reason why we have, say, Steam for Linux. The company that understood the need to own the whole stack. And the company that couldn't sustain its dream running on a broken, fragmented foundation.
The death of Unity will allow Canonical to re-focus, and hopefully infuse fresh breath into the stagnated desktop. Push the regressions away and make something good and practical that people can actually use and love. I really hope so, because I don't want Google on my desktop. If you think Windows 10 is an insult to your intelligence, just wait. For all our sake, hope and pray that Canonical succeeds, otherwise we're going to be balls deep in Idiocracy. You've been warned.