- cross-posted to:
- [email protected]
- cross-posted to:
- [email protected]
I don’t want them to, but I don’t get why Microsoft don’t just roll out updates the way Android/Linux do. Free. They’ve already adopted the Google method of moneymaking by embedding spyware into everything in the OS. Surely they could stop most of their headaches by doing that? Not that I care. Linux user here for roughly 16 years, sometimes having to go back to Windows hurts. It hurt a lot recently, having to dig out an old windows laptop to ironically install Ubuntu touch on to a phone (OnePlus Nord N10). I had tons of updates and had to install features and so much other stuff just to use it. 🤕
Honestly I wasn’t expecting to see a significant increase. It’s a nice surprise. If we get to 6-7% Linux will be much harder to ignore for software and hardware providers. Personally I’m good on software but better wifi drivers would be welcomed.
It is already the year of the Linux Desktop for me and has been for years. Anyone who run GNU+Linux exclusively is already there.
Every year is the year of the Linux desktop
Me too, since 2016. I dabbled in dual-booting on and off since about 2k, but went full Linux in 2016 and have never looked back.
Living in Finland is doubly humiliating. The country where Linux started, and IT chiefs almost in every company are going “oh yeah? I’m gonna use Microsoft products even harder”
I don’t like to say this about people I don’t know but fucking idiots, man.
Europe breaks their own procurement laws to choose Windows, because they are idiots.
Nice to see the word unprecedented before some good news!
best headline of 2025
I mean, just to be clear on what
…
That’d be the red line, there. Assuming you take Statcounter numbers at face value, even.
Incidentally, how have the MacOS and OSX not converged more, speaking of end of life stuff?
Also note the drop in Chrome OS mirrors the rise in Linux so I wouldn’t rule out this just being user agent changes.
Statcounter is… a valid proxy, but I don’t know if I trust it for fine grain changes. Big trends, maybe. Windows overall certainly seems to have lost some ground over time. Whether that’s desktop PCs becoming less popular, the laptop market moving a bit towards Mac, the handheld market being weirdly represented because this only counts devices used for web browsers or whatever else is harder to parse.
Why do they even have two lines for OS X and macOS? It’s the same thing.
They officially changed the name in 2016. I have to assume they changed what the browser reports earlier this year. For tracking purposes it would probably make more sense to count the ARM version separately from the Intel version, but I don’t know if that’s correlated to this change or what is. I’m not a MacOS user, though, so maybe the change is public knowledge and I just don’t know about it.
statcounter checks user agent strings. it probably depends on the mac browser used to report the user agent string.
I know. Then they process those user agent strings to decide what OS it is. The question is why are they treating OSX and macOS as different OSes when they are the same? It was literally just a rebrand.
Thanks for the analysis!
I’m guessing that isn’t the only reason.
Just switched to Fedora Kinoite and I’m loving the immutability of the OS. Oh bad update? lol rollback and golden. Can’t do that shit with each bloatware force push from MS
I mean, good thing, too. They just broke the lock screen this week. That was an annoying half an hour of troubleshooting to do in the middle of a work day.
Pros and cons in each, I suppose.
They just broke
Who did? I didn’t quite follow.
A KDE update broke its lock screen. Locking the computer would bring up a message reading “The lock screen is broken and doesn’t work anymore, to unlock the computer, hit Ctrl+Alt+F1, login and enter this command.”
Interesting that there’s a “failsafe” like that though.
Hah. Those used to be a lot more common, this is actually a bit of a blast from the past.
Early graphical Linux interfaces were constantly breaking and telling you to go back to a terminal to restart your X server or fix whatever was broken manually.
That used to be the “but you have to use the terminal” of very early Linux, back when “can you install Debian from scratch?” was the old “can you install Arch from scratch?”
Man, I’m old.
Yeah, I don’t know who made the offending changes. I don’t think it’s fixed yet, I’m using a workaround at the moment.
It was a ufortunate situation as it coincided with Fedora infra move which lead to delays all around. Anyhoo, the new version, with fix for broken lockscreen on Plasma is available with qt6-qtwayland-6.9.1-3
I know the fix was up for testing, but I saw some people complaining about other issues with it and it wasn’t rolled into the latest live update for me last I checked, so now I’m using a lock screen wallpaper that doesn’t break and I’m not sure I have a way to tell when it’s fixed other than manually checking.
Also, the error message suggesting a way to manually unlock using keyboard shortcuts to a virtual terminal does not match the defaults on my distro, so that added to the confusion.
Say what you will about Windows, but it was a stark reminder of the places where a single monolithic commercial owner would prevent some issues that can happen in Linux/open source projects. A commercial software developer would almost certainly not have shipped something broken in this way, and if they did they would have rolled it back in an update immediately. They also wouldn’t have had a black screen with some tips on how to bypass the issue, presumably, and if they did they certainly wouldn’t have been just… wrong, or mismatched.
Like I said, pros and cons, but it was a disappointing experience. Mostly because… well, yeah, I can understand what happened and troubleshoot it, but a) I didn’t have the time, so I certainly was glad I am dual booting and could just flip to Windows for the time being, and b) a whole bunch of people would not have been able to troubleshoot this or comfortable tryign to do so even if the provided instructions in the workaround were accurate to their system.
And cloudstike did never happen…
I have Timeshift setup for fast and easy rollback.
Bad updates will happen to any OS. It is about if you want to update fast and having a rollback plan. There are dostros that updates more slow as they actually do their own testing and hence was not affected at all.
For me the lock screen instructions worked fine. The kernel update only broke Ghostty but a workaround fix was available the same day. I could also just have booted into the older kernel from Grub.
It did happen, but that wasn’t an OS update, it was a third party update that bricked the OS. The fact that it could do that exposed some Windows practices that are a bigger deal than Linux’s general jankyness when they happen, but they also surface less often for end users.
I thought this particular boo-boo was revelatory because Linux is relatively on the ball anticipating updates breaking the system entirely (one wonders if it should have to be, but whatever). But this was a widespread but specific issue within a random system component. Without googling for it an end user wouldn’t immediately understand what’s going on, and even then there was a fair amount of confusion for at least a day. There wasn’t “a workaround”, there were serveral, as normies and newer users struggled to understand what had broken and how to fix it, and people weren’t very clear in reporting what worked and what didn’t. This all happened within forums and bug reports, with no central source of information or even a centralized official organization informing of the status. Definitely not how that would have played out in a commercial environment, for better and worse.
Also, this is a slight tangent, but can I flag a couple of frequent Linux community behaviors you’re engaging in here that I wish we would get rid of?
One, “it works in my machine” is a meaningless statement. It adds nothing to the conversation and it doesn’t mean the issue is less important. It works on your machine, your version or your distro but not in others. That is every bug, it adds no useful information. In this case, a static screen contains specific instructions that report a common default but don’t match implementation on every distro, so this warning screen isn’t always accurate. That, in itself, is a problem.
Two, “here’s all the smart stuff I did to fix it” (or the smart stuff I do to prevent it) is also entirely useless. The issue came and went, everybody fixed it. The goal isn’t to work around the OS or the DE’s jankiness, it is to have it not be janky in the first place. Putting the onus on the user to fix the shortcomings of the product is… a mitigation, I guess, but the goal is to compete with the paid alternatives on a mass scale, which has different requirements. Complaints about a wonky area of Linux shouldn’t be dismissed or excused with offers to teach people manual workarounds or even best practices, they should be addressed with fixes from the developers of the components that have issues.
Maintainers, I guess, as in, the update that was rolled out, was broken for some users. But I don’t know if that’s the case here
I meant MS or KDE.