Oh shit, is this it? Is this actually the year of the Linux desktop?

This graphic: fuck BSD in particular

In this context what is the difference between macOS and OS X? I can’t help but feel that is like listing Windows 10 separately.

I'm so excited now that I have time again to start tinkering with Linux ✌️

The spike likely represents Statcounter being too slow to adjust their methods of filtering out bot page views after some brand new web scrapers appeared on the scene hoovering up all the data for AI training. 10% linux market share would be a cosmic event. You would hear about it in mainstream media if it were really happening.

Edit: I forgot about popular youtube channel Linus Tech Tips' recent videos in which they gave linux their highest praise yet. One was published in mid June, another in mid July, and their most recent and most favorable to linux was published on July 22 ("My Experience Using the Steam Machine as a Desktop PC"). There have also been videos on the Framework laptops. Vibes tell me that there's a trend of more people using mobile over desktop, so I would expect a higher proportion of desktop users to be the gamer or technically inclined type, enough to be influenced by LTT and give linux a shot, compared to 10 years ago.

Here's the data for the last 4 months.

   Date    Windows     OS X    Linux    Chrome OS    macOS
2026-04      66.39    17.04     2.94         4.42     9.21
2026-05      64.66    19.92     3.56         3.72     8.15
2026-06      59.08    24.06     6.14         2.19     8.53
2026-07      57.63    21.07    10.61         2.07     8.63

The data looks noisy. You would expect the lines to have consistent, regular annual patterns. You can see that with Chrome OS which seems to have a dip in the middle of the year corresponding to when schools are out for summer vacation, and if you scan upward, it appears Windows has a corresponding hump at that time. But there are still unexplained extreme swings. The lines look smooth up until 2012/2013. Maybe Statcounter doesn't have as many members/page views as it once did, leading to higher variance. We could see that if we had absolute numbers, but we only have percentages. That's probably affected by the advent of the LLMs. Nobody's going to the web anymore to find information.

There's no way that's accurate, right?

IIRC this was debunked some time ago, it's related to a change in OSX/MacOS that's causing Apple devices to show up as Linux.

It may not look like a big deal but it wasn't too long ago Windows was like 95% of the OS market.

Next year, I'm telling you, it's going to be the year of the BSD desktop.

The MKBHD effect?