Updated: October 22, 2012
There's a silly sentence. I am projecting what I feel ought to be true, like so many other bloggers out there, except that I am neither a blogger, nor wrong. The thing is, I want to talk to you about where the Linux desktop, I repeat desktop, is heading in the coming few years. A hunch, based on years of being absolutely right on everything. Read my past articles, and thou shalt be convinced utterly.
So what now? All right. In 2011 or so, more or less, the Linux world suffered a split, with the Gnome 2 desktop being torn apart and replaced with a mutation called Gnome 3. This led to much dissatisfaction in the community, and two alternative projects were born, one called Cinnamon, the other MATE. The former uses the Gnome 3 technology, but makes it more presentable and usable, while the latter is Gnome 2 reincarnated. The Pauli exclusion principle tells us you cannot have so many desktops around. So which one will survive into the next decade?
There's more than just Gnome being chopped apart. You also have Unity, which adorns Ubuntu, and this one also uses the Mk.III libraries in its core. Now, given the rapid market growth of Ubuntu, not very likely to go away. Steam being ported to Linux, Ubuntu is the strongest indication that good times are ahead.
KDE is another one, a steady workhorse of the Linux world, moving through its incremental versions like a champ. At the moment, it has reached version 4.9, which is quite stable and highly usable overall. However, from the growth perspective, KDE has always been a stagnant project, because it was used on distributions that did not experience any expansion in their user base in the past decade, and because it targets its very own segment, which does not really overlap or transgress into the world of other desktop environments. A workhorse, if you will.
As for the rest, like Xfce, lots of good stuff happening there too, but these are community efforts that cannot scale and compare to the big guns in this league, so we will not debate about them at this time.
All right. The way it seems, to me, KDE will remain a highly popular, conservative choice with a very Windows-like mindset and workflow, so it could be a good potential candidate for Windows refugees, especially those disillusioned by Windows 8. Unity will become the trademark of what Linux desktop looks like, and maybe mobile too, because its growth will go hand in hand with Mark's aggressive and seemingly largely successful business campaign.
Which leaves us with Gnome 3, which no one really needs, and the two replacements, which also no one really needs. Wait, don't go mad. Yet. Wait and read till the end. Anyhow, no one really needs, I said, and that's how it seems. Because, if you think about it sensibly, Gnome 3 did not usurp Gnome 2. It came as a replacement, and failed, but the real reason why Gnome 2 slid away into the sidelines was because the largest Linux home distro decided to switch to its own home made environment.
The sidelining of Gnome 2 did not go quietly. People clamored for a desktop environment that is simple and pleasing to use. Part of it is habit, part refusal to change, part pure logic. Because, if you judge Gnome 2 from the efficiency angle, it remains, by far, the most productive environment created in the last two decades. It offers the combination of visibility, access and simplicity that are not matched by any of the rivals.
And so, high hopes were invested in Gnome 3, which turned out to be a poor sibling of Unity or near as it makes no difference, with even less usability and logic all around. Meanwhile, Unity has slowed progressed to a point where it actually makes sense for most people, and even geeks like me can use it with only a slight degree of discomfort. A lesser evil, if you will. However, Gnome 3 remains just as unfriendly as on its first day. Which means that all those hopes for an upgraded Gnome 2 successor were dashed into tiny pieces. This in turn created a vacuum.
The vacuum was exploited by the Linux Mint team, which enjoys a highly enviable strategic position in the Linux world. Forking off Ubuntu and with success, unlike most other distros out there, it can afford to take the good stuff from its parent and discard or improve the rest, thus making good better. Which is exactly what the team has been doing, translating directly into user base and popularity. Because of its semi-flexible work model, where they embraced some of the new stuff that Ubuntu offers, blended with a conservative and efficient approach to productivity, Linux Mint is now the most popular distro out there.
Part of the reason is Cinnamon, which is a modern version of what Gnome 2 would have been if transformed into Gnome 3 the right way, and this while using the newer Gnome 3 technology, on top of the solid and growing Ubuntu base. The best part is, the layered approach allows people to run Cinnamon alongside Unity and other desktop environments, so nothing is lost in fact. There's no need for compromise, and even Ubuntu looks great with Cinnamon. Hell, even Fedora does!
This way, the Gnome 3 presentation layer, which sucks mega turbo level, is taken out of the equation, the technological development is maintained, and whatever success Ubuntu reaps will translate into Linux Mint success. Unless Canonical decides to be greedy and cut off their friends, which it most likely will never do, you end up with two Ubuntus, the stock one geared toward more flexible-minded users, the other designed for classic desktop users, which form the majority of all computer users worldwide. For people with the youthful Android-like attitude, these new form factors and UI logics will appeal. Mac people might start doubting their religion. And Windows people, slowly becoming more and more sick of the Windows 8 nonsense, will start flocking to Mint, which offers a natural continuation of what they know and love, without any loss of modern and cool and efficient. This way, the entire spectrum is covered.
One might argue that you do not need MATE if you have Cinnamon, and vice versa. Not quite true. MATE is still the undefeated champion of low-end resource usage. Machines that cannot run Cinnamon, or Unity for that matter, and no more Unity 2D starting with Ubuntu 12.10, will putter along just fine using MATE, because MATE can run without any visual effects whatsoever, even on pure VESA if you want. And this desktop environment still remains the best example of KISS. Quite unbeatable.
MATE remains Gnome 2 in looks and functionality, and this is not a bad layout. MATE offers the practicality of reason, plus it can infuse semi-failing distributions with a new sense of purpose and use.
Yes, but it's not a magic wand. Cinnamon 2D merely moves its rendering from the graphics card into the software stack, which means your CPU is responsible for drawing objects on the screen. This is an intensive task, which leads to a high processor usage, a big penalty on old, weak hardware. It's a paradox really. If your machine is powerful enough to run Cinnamon 2D without any apparent performance impact, it probably comes with a handful of revvy cores and whatnot, and this means you can mostly likely use Cinnamon just as well.
This is how it goes. Ubuntu will become more and more visible, and even ordinary people will start hearing about it. They might even begin using it. They will be pleased by the free and quite lovely operating system, although for most, it will be an app-and-game shell, nothing more, and that's fine.
Then, those seeking productivity will naturally converge into their comfort zone, which calls for a traditional desktop. Yes, yes, while you may say that 30 years from now, when the PC dinosaurs die off, everyone will be willing to work any which way, you cannot deny the simple human logic of keyboard and mouse and where things should be, no more than you can deny QWERTY dating back 150 years into the past. Thumbs are thumbs and not getting smaller, no matter how touch-sensitive your screen is.
The traditional desktop is also the natural escape for Windows users, which comprise the vast majority of consumers. When you combine the two, classic desktop + Windows, you get an almost perfect overlap, hence Cinnamon and MATE will be the familiar thing that everyone wants. And when the first taste of cool wears off, practicality will win over. The beautiful thing is, you will not need to install an entire operating system, just a single package. Therefore, desktops become apps. Colorful apps. But not in the Metro sense, no. In the sense that you can dictate your productivity, not the other way around. So apps, it is. And everyone can find and install those, especially when they use an app-market-like program. For free, no less.
This way, if you feel like it, you can run Unity on Monday, Cinnamon on Friday. You can actually combine productivity and cool any which way you want. Your partitions remain unchanged. Your data is untouched. And if you go for an OEM solution, you're safe, too.
To sum it up, Cinnamon is the more practical mirror of Unity. Same underneath, so very much different above the surface, fully inclusive. MATE is the same mirror, only more suitable for low-end machines and hardcore functionality lovers. A harder sell, but in emerging markets, where people cannot afford USD1,600 to buy a toy, the ability to get the maximum from existing available devices will be much appreciated. More importantly, emerging markets constitute half the world or more, so you cannot dismiss those just because they are uncool or do not live in the San Francisco area code. However, MATE will not conquer the world, for the same reason Gnome 2 did not conquer the world, and not because of the desktop framework, but because of the distributions shipping with it, the same problem that has affected KDE for the past decade or so.
You might be a little confused with what we have here, so let me help you. The ability to install multiple desktop environment has always existed in Linux, but there was no need to exercise it, because Gnome 2 and KDE always catered to entirely different segments. Unity and Gnome 3 broke the barrier. Their alternative solutions are in fact highly compatible add-on layers rather than complete replacements.
The same way Mint builds on Ubuntu, so do Cinnamon and MATE build on the existing frameworks, expanding the potential user base. People who find Unity only 70% good can make it perfect with Cinnamon. Nothing is lost. You don't even have to abandon Ubuntu. People who want Gnome 2, get it. People who hate Gnome 3 now get a perfectly viable solution. It all comes together.
My prediction is, Ubuntu and Mint will grow side by side, covering a much larger market segment than they could have otherwise. And because Cinnamon is such a darn simple extension, so to speak, it will eclipse, ever so slightly or more, its parent. You must never forget that Windows people will not take lightly to a left-positioned launcher, but they will almost naturally flock to something that looks like Cinnamon. Forget about the geeks, even though they are the reason why it was born in the first place. And it seems to be working, as you can see the Mint popularity soaring. Something is being done right somewhere.
MATE will follow, to a much lesser extent, but to all those who find Cinnamon too bloated, too heavy, incompatible, or simply do not wish to use Unity, Ubuntu or alike, it will be a great supplement for a productive and efficient computing. However, it will NOT make a revolution of its own, because Gnome 2 failed in the same space. So it must rely on Cinnamon for some brotherly love.
And so, the Linux desktop world will evolve to a new level, where only KDE will be the familiar player from years gone. You might think that too much forking causes strife and leads to significant fragmentation, and in most cases, this is true. But you cannot treat Cinnamon or MATE as separate pieces, which is what makes them so useful. Hence, the reason why they will emerge as dominant desktops in the years to come. The future of anything is something that looks like the present.