Linux Desktop Growth Trajectory

Is it finally the year of the Linux desktop?

9 May 2026

Linux has been making massive strides lately, so why not do some growth analysis on it?

Interested on some background and analysis of structural incentives for Linux’s growth?

There is some real competition in the OS space now.

However, I don’t want to talk about subjective perspectives. For example, someone might actually like forced updates. I don’t know. I’m not a masochist.

Rather than convincing individuals to change their preferences, I want to analyze systematic structures that can have a broader impact on how people interact with their computers.

First off, an operating system is not profitable. You have to constantly fix it, maintain it, and add new features. Users constantly expect it. There’s a reason that planned obsolescence is a thing for Android and iOS. Even if Windows 10 was supposed to be the last version, Windows 11 was launched because of this (now you know why TPM is an arbitrary requirement). However, as a loss leader, an operating system is the gateway drug to sell you subscription services. Microsoft being the dominant player has constantly pushed you towards OneDrive, Office 365, and now AI. Apple will eventually do the same, if they aren’t already. Apple may not seem like it’s enshittifying today, as it’s currently subsidizing costs to attract users frustrated with Windows into its ecosystem, but one day its shareholders and partners will come knocking.

Next, computer manufacturers are trying to respond to the MacBook Neo. Some YouTubers have pointed out that some hardware manufacturers haven’t given them products because they did not want to be compared to the MacBook Neo. Windows has a $100 license fee. Mac OS only exists for Apple. What’s left? Linux. Companies can even customize Linux to cater to their users’ needs, if needed.

On the gaming front, Valve is actively trying to push gaming on Linux. When Windows 8 first launched, Valve clearly saw the new Windows store as a threat to their business model. Apple’s app store was even more restricting, so Linux it was. Fast forward 10 years, and Steam hardware is rivaling traditional consoles. I’m honestly quite amazed at Valve’s engagement with the FOSS community, but financially it makes sense. Valve has no idea how to make an OS, so funding FOSS developers was the next best thing. FOSS developers are very passionate about their project after all!

In Android land, Google has also announced that it will be merging Chrome OS into Android. A major shift as Google is trying to use Android’s popularity to get into the desktop market. It has always had its sight on the desktop market ever since it launched Chrome OS, but was never truly able to penetrate deeply due to Microsoft’s dominance. My guess is that Google will push for an Android phone to also be your laptop/desktop. Basically a 3-in-1 device. In the future, other companies may also expand into a 6-in-1 device where it becomes a phone, laptop, desktop, smartwatch, VR headset, and a tablet. But that’s just speculation based on the progress of how foldables are going.

In geopolitics, Linux and other FOSS projects are making inroads in Europe. Unless something major happens, Trump will be the president for three more years. That means at least three more years of hating on Europe. American tech companies will cater to Trump’s attitudes and Europe will have to respond with digital sovereignty. And there’s two approaches that are currently happening. One, creating their own big tech solutions, which personally I don’t think is the correct answer as that’s how the original mess happened. And two, it is to rely on FOSS solutions like Linux.

Finally, there is no one major player dominating Linux. People have constantly complained about the sheer amount of variety in Linux distributions. But that’s its strength. The GPL license, in contrast to permissive licenses like BSD and MIT, will ensure that Linux is always decentralized in nature. Don’t like Ubuntu adding AI? Switch to Arch. Don’t like Arch’s verbosity? Switch to Linux Mint. Hate how Linux Mint is too outdated? Switch to Alpine. The world’s your oyster. You actually have a choice and no one company/individual can take that away.

These are all reasons why I think Linux will grow.

Now that I’ve set the stage on Linux’s trend, I want to see how it’d look in the future. It seems that the best way to model growth is something called the sigmoid function. In short, growth is initially exponential, and at a certain point, called x0 (pronounced “x naught”. Supposed to be x subscript 0.), it starts to slow down. Zoom out and the overall shape will look like an S.

There are some things that could turn this trend around. One is for Microsoft to stop shooting themselves in the foot, but that will never happen because of the amount of money being turned into AI. Another way is for Microsoft to turn Windows into a Linux distribution. Yep, you heard that right. Again, the operating system is not profitable. Someone has to maintain it, someone has to fix the bugs, and someone has to cater to the customers that complain. Linux already does that for free (as in beer). Microsoft can then continue to sell services. However, if Windows ever does turn into a Linux distribution, I highly doubt it would be considered one. Just like Android, it probably won’t contribute to the wider ecosystem and will always look to further control the narrative. I mean, companies are designed to rake in as much money as possible and the best way to do that is by siloing their userbase. And yes, if you were wondering about Android translation layers for Linux, there are a couple out there: Android translation layer and Waydroid. If Android were truly Linux, there would be no compatibility layer.

Why Companies Enshitify

At a high level, companies “enshittify” because of structural incentives. It’s a natural consequence of trying to maximize profit. That sounds obvious, but it helps to think about it more like an ecosystem than a moral failing. Different systems behave according to their underlying incentives.

In this ecosystem, I see two opposing forces. One system is driven to continuously accumulate resources. The other is driven to minimize effort and make the most of what already exists. In software, companies tend to align with the first: they constantly ship new features, even if those features create new problems that can later be monetized. Free and open source software (FOSS), in contrast, often reflects the second: doing as much as possible with as little as possible.

In history, we’ve seen this play out repeatedly. Projects like KHTML quietly shaped modern browsers, and BSD heavily influenced systems like macOS. Yet the pattern is consistent: permissively licensed work gets absorbed into proprietary ecosystems, and the original contributors rarely benefit proportionally. The greedy system always takes, integrates, and builds closed layers on top. And that’s whats rewarded.

That’s why licensing matters. Permissive licenses (like BSD or MIT) allows anyone to take open work and close it off to others. Copyleft licenses (like GPL or MPL) take a different approach: they restrict that specific freedom in order to preserve the right for anyone to view, modify, install, and run the software(side note: Linus Torvalds would like to disagree here, but that’s what the license states. Unfortunately Linus is single handedly allowing planned obsolescence to occur, instead of forcing hardware manufacturers to allow users to unlock their Android bootloaders.). By requiring derivatives to remain open, they prevent the original work from being absorbed and locked away.

This is the “paradox of tolerance” in software form.

Systems create behaviors and repeated behaviors create morals. What system would you like to create to get behaviors and morals that you would like to see? I hope to generally explore this question more in a future blogpost.

Graphs

Note:

Steam

Steam

The purple line shows the overall Linux usage percentage on the Steam platform. The black line represents Steam’s Linux usage percentage among English‑speaking users.

Since this is released monthly, Linux is projected to reach 50% marketshare in 200 months for the English speaking side of Valve and 243 months overall.

In other words, Linux will be 50% marketshare in 9 years for the English speaking Steam population and 12.5 years overall.

Pornhub

(Yes, I did make the header a bit bigger. So people looking at your screen could definitely know what you’re looking at. XD)

Pornhub

For Pornhub, Linux will reach 50% in 40 years after data collection started. 11 years have already passed, so Linux will reach 50% in 29 years! Yikes!

Why the discrepancy? I’m guessing the data collection before 2018 was flawed. It seems strange to me that 1/3 of Linux users disappeared off the face of the earth or that there was a mass influx of users buying Windows and Macs between 2017 and 2018.

If we only account for data from 2018 onwards, then Linux would reach 50% in 2038. But arbitrarily removing data seems weird. If Linux truly has the momentum, we should see it reflected in the graph in the coming years.

Penguins, time to assemble o7.

(And here’s another shameful text so other people glancing at your screen would do a double take. XD)

Pornhub

USA Gov

USA Gov

I’d argue this is the most interesting graph, as we have DAILY data points spanning all USA government websites from 2018! Let’s take a quick peek, shall we?

Linux is projected to hit 50% in 7138 days. 3042 days have already passed, so 7138 - 3042 = 4096 days left to go. This roughly translates to a little over 11 years.

Note: They seem to have switched analytics partners in the middle. v1.1 covers 2018-01-01 to 2024-06-23 (Universal Analytics). v2 covers 2023-08-01 to present (GA4). I used Universal Analytics until it’s cutoff month, and then switched to Google Analytics for the months after.

Slight bugs in my analysis:

Analysis

From these graphs above, it really seems that Linux will be a dominant OS within the next 10 years. If trends change, then maybe this course will also change, but it seems unlikely. I’m hoping Linux makes a lot of progress before the wind shifts.

Recap:

If current trends continue, I expect the following to happen:

Interested in Switching to Linux?

For most people, I’d recommend just two Linux distros:

(And this is coming from an Alpine Linux fanboy.)

If you want to explore other distros, use Ventoy to load multiple ISOs onto a single USB drive and try them out in a live environment. Most distros are more similar than they seem. You may need a bit more elbow grease depending on the choice.

Why these two?

It’d be nice if distros also make a simple executable to try/install Linux, instead of using a flash drive.

You Don’t Have to Commit Right Away

There are many useful guides on how to switch online. I would recommend YouTube as a better source than Google for how to pick and install a distro. Even Microsoft has a guide.

You can always go back to your current setup if you don’t like Linux. And if you’re worried that you’ll mess something up, just make a backup first (which is good practice anyway).

Contact Me

Questions, comments, concerns? Please reach out to me on my Mastodon account. I’m quite active there and can follow up. If you have any other websites I can track or have any analysis that you think is insightful, please do reach out. I’d love to hear from you.