Instead, I think it's that "Linux" is an overloaded term. One sense is that someone downloads and installs a "Linux" distro because they actually want to use "GNU+Linux" (wink). The other sense, what you are aluding, to is that linux is foundational to most things IT. If I subscribe to DSL the provider is probably going to send me a modem that runs linux. But that doesn't mean I chose linux. I just wanted DSL. Same for Android. Most people that use Android didn't choose a linux-based mobile operating system. They want Android or are just using whats one the phone they wanted. And indeed, I don't think many "GNU+Linux" people would tolerate the specific essense of Android in their distros.
Now, SteamOS might be the bridge between these two worlds. On the one hand, Steam Deck users also didn't chose Linux. But then, the resources that Valve can spend on enabling gaming on Linux because of the success of Steam Deck means that many more people, like me, can finally consider choosing Linux.
In some comparisons it makes sense to remove DSL modems from the equation and focusing on some more "computer-like" subset of device (otherwise no OS would ever do better than "several orders of magnitude fewer deployments than Linux"), but discounting every device where you did not actively pick the OS would make for an extremely biased comparison.
Plus, we don't care about such distinctions outside comparisons. SteamOS have already driven significant improvements for regular desktop users. Same goes for Tizen, Android and ChromeOS. Even wonky DSL firmwares have positive effects for the rest of us.
> GNU+Linux
no.
For many non tech savvy users if you give them a Linux laptop with any modern DE, they can't tell the difference.
But it seems these filters are almost always cherry picked against Linux for some reason I sincerely don't know.
In that case 99.9% of windows users fall in the same category. They are just using the OS that came on the device they bought. How many people actually bought a separate license, and downloaded a Windows installer because they specifically wanted to install Windows on their device?
Pretty much the entire (DIY) PC gaming market? Granted they may not buy Windows but they are certainly making the choice to download and use it over linux.
Sure there is a pre-built market for desktop PCs but I would think the PC Gaming market skews to DIY though I have no stats.
Note: I run arch btw.
Having said this, the local library from my German city does run on GNU/Linux.
I'm not sure what would make you an authoritative source on why people by Macs or where you're sourcing your thoroughly researched data, but I can tell you I purchased an MBA to run macOS and the software that runs on macOS. Or like the Intel N100 mini PC I purchased explicitly to run proxmox + OPNSense. Or like my Windows laptop to have a mobile lab.
I pick the machine based on the software I want to run. I'm sure we can find one, perhaps if we stretch it, two other people on HN who also purchase machines based off of the OS/software they need to run.
And the only thing he use it for entertainment is browser (for reading novels)
If I am managed to setup all these properly on other system. I doubt him would ever notice a difference.
Non tech user just can't care about what system he is using less. All matter for him are tools that help him to finish his work.
What about dvd/cd or install some new softwares? Well, I am the in house tech support.
Most users do not watch WWDC and do not know what OS release notes are. They don't know what the boundary between their web browser and their OS is, and macOS just becomes "the thing that nags them to update it" - a nag that users unfortunately still ignore, as evident by my recent confrontation with Big Sur machines.
That was a majority use-case between 2005-2015. Now it's a niche case (yeah, even if absolute numbers are still big enough, relatively speaking it's a niche)
If Linux wins in that category as well, could we finally stop moving the goalposts?
However, that doesn't mean that every non-HN user out there buys the aluminum-shell-in-the-vague-shape-of-a-laptop and doesn't make a conscious choice.
My point still stands. The GP doesn't have the data to back up the assertion.
Most desktop computers still do, and in alternative there are USB stick variants.
Which doesn't change the fact regarding what drivers, for what set of OSes, are supported out of the box.
For good reason. Running those updates requires restarting the computer, which in turn means you have to get your working state setup again after each update. People use computers to get things done, and running updates often has the opposite effect.
> That does not mean that every non-HN user...
Not every - but I do believe it is still the majority. There are a lot of people out there, and the knowledge required to make the aforementioned OS choice is niche.
This is not implying stupidity, just that there are many trades and interests out there, and subscribing to ours specifically is not a given.
Just a few weeks ago, I tried playing a Japanese game that I was interested in, but it didn't even get past the start menu. Not a big deal for me as I'm not an avid gamer, I just moved on.
But imagine telling a hardcore gamer that no, you can't play that new game released an hour ago. Maybe it'll become compatible a few months later, maybe not.
Or tell a professional artist that no, they can't use that art software anymore, they have to retrain their entire skillset with an entirely different software.
Now that's a show-stopper. For most people, it's 100% compatibility or bust.
It is true that if you try to reuse your 2003 Office install you will fail, but such unsupported and deprecated software will also cause trouble on a modern PC. Even if it runs, using an old version of Outlook is extremely unsafe...
Common end-user hardware does not require drivers nowadays either (not even printers due to IPPAnywhere and co., even though manufacturers still ship them for some reason).
Things are always hairy outside that though - macOS no longer permit kernel extensions, and you know you are in a dark place if DKMS gets involved on Linux.
It's VERY EXTREMELY IMPORTANT to "teach the controversy" of GNU+Linux. It's not at all important to actually use the term.
Which is why I made no assertion with a percentage.
> but I do believe it is still the majority
What makes you believe that? Thousands of data points? Hundreds? Tens? The handful of people around you? Certainly not statistics.
Even if things continue to work, updates may come with UI changes that disrupt the user's workflow (and the trend tends to be to remove or hide functionality and make things less useful/information dense as time goes by). Unless there's something you specifically want, modern software updates tend to be high risk low reward from the user's perspective.
Incidentally this is why I prefer desktop Linux. With the exception of Firefox turning into Chrome over the years, FOSS software tends to be remarkably stable. My home computer feels like it hasn't changed at all in almost a decade. I never have to fiddle with it.
Professional artists are also an interesting choice. If they are in a major art studio, they are probably having to use custom tooling there already. And Wacom is very well supported on linux.
Ridiculous standards aside, I find it an extremely reasonable to assert that given more than 7 billion people across vast areas, interests, ideologies and jobs, and given the vast hi.an knowledge, expertise and culture, that any particular interest or knowledge is only shared by a small subset.
That's not at all what they said. All they're saying -- correctly -- is that you are asserting some sort of magnitude ("majority") without any data to back it up.
I would guess, though, from my personal experience, that you are probably right that a majority of people just get whatever laptop with whatever OS they're used to because that's what their parents/school/employer gave them to use, and when it's time for a new machine, they just get whatever they had before. But I don't think this is a very large majority.
> I find it an extremely reasonable to assert that given more than 7 billion people [...] that any particular interest or knowledge is only shared by a small subset.
Probably true, but also remember that OS choice isn't always driven by interest. A macOS user may get frustrated with the state of gaming on the Mac, and decide to switch to Windows. Or a Windows user might really want or need to use an application only available on macOS (though I expect this sort of thing doesn't happen as often anymore, since more and more of people's computer use ends up being through a web browser). A Windows user might also buy an iPhone or iPad and get into the Apple ecosystem enough that they decide to switch to macOS.
Certainly some people who do switch OSes don't do so because they've made an independent choice; they do so because they switch employers, and something else is the only thing available, or a friend evangelizes another OS to them to the point they want to give it a try, and end up liking it.
Regardless, many people these days don't even have a laptop or desktop computer, and do all their computing on their phone or tablet. I think that 7 billion number gets a lot smaller when you consider that. (Also, as an aside, the current world population is estimated to be a bit over 8 billion now, not 7.)
I have a game that works fine with proton on one machine, but doesn't on another machine. But works fine with windows on both machines.
I tried for hours to get it to work on the other machine, since it clearly can. No luck.
In the end with proprietary software, it's always a black box.
This is also an odd rabbit hole to fight about. I've had so many windows boxes that couldn't play games throughout the years that I've basically accepted that games programming is hard. :D
No difference of being a nice package for something like MAME.
WWDC sessions on that roadmap state quite clearly that is the long term end goal, all the kernel extension mechanisms will only be available in userspace, with one year transition for each subsystem after an userspace API is made available.
Exactly. The thing many users are referring to as MacOS, is in fact, Darwin/MacOS, or as I've recently taken to calling it, Darwin plus MacOS. MacOS is not an operating system unto itself, but rather another component of a fully functioning operating system made useful by the BSD corelibs, shell utilities and vital system components comprising a full OS as defined by the Single Unix Specification v3.
Your response is more reasonable, and is also more out in the open about the alternate belief (of course equally without data). Nothing wrong with disagreeing - only about making up unbalanced burdens of proof in casual discussions.
My biggest issue with Windows was the forced reboots.
I don't see why. I really like macOS. More than any other environment it's uncluttered, well organised, it gets out of the way. It abstracts a lot of unnecessary technical detail but still allows diving into that, if I'm inclined. And most of all, I like well integrated system-wide features instead of palming it off to developers. Sure the hardware is good too, but I don't know where you get that idea from.
A modem's application is the one thing it does and the interface is an abstraction layer standardised enough that the modem could be replaced in a shoe closet without anyone even noticing.