There's such a deep seeded, systemic bias against linux that it actually can never win, to any degree or magnitude, because the moment it starts winning we just move the goal-posts for the flimsiest of reasons to ensure it can't quite claim that victory.
Linux is obviously and clearly the most popular operating system kernel on the planet. Oh, no, that's no good a measure, servers are messy, let's refine it to most popular consumer operating system kernel? Oh... it, could also reasonably claim that title? No no, no Android, that doesn't count. Nope, No Chrome OS either, you can't have that, that's, well, that is linux, but its not. Just nice, pure, desktop linux, yes, perfect, arch linux, kde desktop, that'll never trend up and thus is the perfect new-new definition of desktop linu--wait hold up, I'm getting word this is, not possible, its actually SteamOS? Nope, kill it, that's not desktop linux either, kill it.
As to your latter point, you're just muddying the topic. Marketshare/netshare metrics are subdivided as they are because most people viewing those reports don't care what their web server, handwatch, video game console, etc use. They care what OS the computers people actually interface with runs; mostly for targeting reasons.
Deep seated.
https://www.merriam-webster.com/grammar/deep-seated-deep-see...
So it is proper Linux, as GP comment implies. Yes it's running games in Windows compatibility layers, but it is also a complete Linux system itself, with desktop. Definitely counts as running Linux.
And a decent chunk of those games are running on the Unity or Unreal runtimes. Do they count as "running on Windows"? Where are we drawing the line here?
excuse my ignorance, but are there any major developer tools that don't support Windows? I can only imagine some internal enterprise tooling doing this.
>Ultimately compatibility layers are good because they reduce developer workload so that developers can focus on what really matters
I don't mind them as a concept, but I personally want as few points of failure between me and my software as possible. Some software is already either overly bloated or buggy (or both) as is without wondering if there's now compatibility layer issues on top of it.
if the developer released it as a windows build but is being played though a compatibility layer, yes. Unity and Unreal both support deploying Linux builds, but it doesn't mean making a proper Linux port is as easy as pressing the "Linux" button.
>Where are we drawing the line here?
I don't personally care for what counts or not. I just personally wish for more native support.
I thought it was relevant information and could predict it was probably steam deck.
The newer generations of users are not computer literate, they use user friendly computers. It was not a conscious decision based on the OS it runs on.
The people that care about OS - specifically linux - dominance are not those users.
At all, or well? Because git works on Windows, but it's meaningfully slow as a result of NT and unix having very different ideas about how filesystems should work.
put it this way: I have the expenses and know how to simply have 2 native platforms if I need it. If I need to emulate something but don't feel like spinning a second boot, it's usually for a few niche programs I need very occasionally. Otherwise, I can just get that OS ready.
That's how I see the Steam Deck. I simply have a portable windows machine that has Steam on it, so I don't see the incremental value of having these games run through Proton while having a linux machine. Because I have "running windows games" covered. If it was running Linux games natively, I may in fact buy it simply to help say to developers that I want more native games on Linux. But if devs just keep focusing on Windows, I have that set already.
And To be honest I'm not a huge fan of the form factor nor screen to begin with, so it'd be a bigger compromise choosing to play on a Steam Deck than if I could choose other machines.
Depends ... It can be that easy, sometimes. I was maintaining a huge Unity based VR setup the last years which had both Linux and Windows PCs (mostly for legacy reasons). Building for both platforms was done from the same bash script with the only difference being the platform identifier.
Tbf that was a very standalone application that did not interact with the OS a lot, but otoh I would assume that a lot of games are like that.
The other is that the notion of a 'proper port' here doesn't really matter. Source ports are rare and don't necessarily turn out better than compatibility layer ports. If the goal is to have playable games, or for Linux to become usable as a desktop OS for gamers, what kinds of ports we get doesn't come into it.
That said, source ports are nice and it'd be nice if they were common some day. It would also better secure desktop Linux's position here.
Probably the reason it can never “win” is because it’s an operating system kernel. It’s software.
You know what can win? People. Companies. In competition with each other. The Linux Foundation wins all the time when it gets contributions of code, money and capital. Valve wins all the time when people buy Steam Decks. Microsoft wins all the time when someone spins up a Linux server instance on Azure. Users when all the time when their Linux systems do the things they want it to.
But yeah, Linux never wins. Never can, never will. The same is true for Windows and Mac OS X and FreeBSD though, and postulating that a piece of an operating system code base can “win” or “lose” is the linguistic trap that sniped millions of nerds for over three decades. There’s no scorekeeper in this game.
Semgrep is one that I use at work. Nix is another. Docker¹ is a third. Many terminal emulators support multiple operating systems, but not Windows.
Windows support also often lags for new programming languages. Golang didn't run on Windows at first. Crystal is only now starting to have full-fledged Windows support. Plus there are many tools that do run on Windows but work poorly or are extremely slow or require tons of compatibility shims, like Git and Emacs.
A lot of dev tools are Unix-first. You just probably use only a few of them if you work at a Microsoft shop.
--
Not Docker itself at this point but 99.9% of all Docker containers that anyone actually uses.
It is what it is. I would say they will never focus on Linux but it's not so clear anymore. A lot of things are changing fast.
> And To be honest I'm not a huge fan of the form factor nor screen to begin with, so it'd be a bigger compromise choosing to play on a Steam Deck than if I could choose other machines.
Luckily even better handhelds seem to be coming out. Asus ROG Ally (windows) is worth a look, as is Ayn Odin 2 (android..Now we have this too as a potential gaming platform to consider. Maybe not today, but who knows.). I recommend ETA PRIME's youtube channel to keep an eye on these things.
I consider 2 things to be the same OS if they can natively run the same unmodified binary files and if we look at it this way, Linux is on a losing streak: - I cannot download an app on Android and run it on Debian - I cannot compile a program on one version of the same distro and run it on another version of the same distro
The moment I will be able to compile a program that runs on both Arch and Android, we can start adding the stats up, however I doubt tis will happen anytime soon looking at the poor attempts at fixing this.
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.
As someone who has played a lot of games with native linux "support", I want less of it.
In nearly every instance of these native ports, switching to using the windows version via Proton was a better experience, either because the Linux version was outdated, unmaintained and buggy, or it simply performed better.
Annoyingly, as far as I can tell, Steam these days doesn't make a distinction between native ports and Proton games so it's hard to tell if I'm getting served the unloved child version until something goes drastically wrong and I have to start messing around with it.
Any kind of port can be of high or low quality, though.
Linux "won" the server. Linux has not "won" the consumer desktop/laptop gaming market.
SteamOS is super interesting. Does it count as Linux? Yes and no? Depends on your goalpost! If I were shipping a game today I would 100% support SteamDeck and SteamOS. I'd maybe provide support via Win32 emulation, or maybe native. I'd probably stick to Win32 and only do native if needed for performance.
But I would 1000% NOT claim to support "Linux". I would support SteamDeck and that's it. If any user reported a bug or issue on other Linux distros then I would close the bug as "not supported". If it works, cool. If it doesn't, that's cool too. You're entirely on your own.
Supporting one Linux distro on one piece of hardware is pretty easy. Supporting all the Linux flavors on all possible hardware configurations is a bloody nightmare. And it's radically harder than supporting Windows across all hardware configurations. And it's definitely not worth it to increase sales by ~0.5% or so.
Take from that what you will.
this is fascinating..
Is it possible to install Arch linux and add the SteamOS layer and cut over and back again as desired?
You can also just go to the normal KDE Plasma environment from the steam menu and use it like normal Linux - or even install let's say emulators from Arch repos and add them to Steam then run them from SteamOS interface.
Funnily enough these goalposts were actively moved by Linux proponents themselves. Because whenever there's a criticism of Linux, it turns out that the system in question is not a true Linux... because reasons.
> Linux is obviously and clearly the most popular operating system kernel on the planet.
It is. I don't think anyone is denying that.
> No no, no Android, that doesn't count. Nope, No Chrome OS either, you can't have that, that's, well, that is linux, but its not.
This has been the argument of Linux proponents (many of them) for years.
> Just nice, pure, desktop linux, yes, perfect, arch linux, kde desktop
I've seen multiple claims that Ubuntu is not the true Linux. From people using Arch, or Gentoo, or...
I personally have not owned a Windows computer in the last 10 years, and even then I only used it for gaming and not for development. If my code works on Windows without a compatibility layer that’s a complete miracle.
Many Windows developers similarly have little idea how Unix works and stay in the Windows development ecosystem.
Ultimately you can only fit so much in your head and I don’t have room for Windows to live in mine too. I’m sure a lot of Windows devs feel the same way about Unix.
It does to me, but I'm not a stranger to being part of an underrepresented niche. I have windows machines to make games playable if I want to play a windows platform game.
I simply prefer control where possible, and leaving something to a compatibility layers makes me rely on two separate platforms being maintained and contributed to (and/or not enshittified) in order to not be SO, be it now or in a future. 3 if you count Steam's contributions on Proton and choosing to carry whatever game you want to play. There can still be bugs in the game proper, but a native port eliminated points of failure for me to investigate.
SteamOS solved the chicken-and-egg problem and is demonstrating that Linux-based gaming is viable. 2% isn't a massive number, but it might just be enough for a game developer to justify compiling a Linux-native version too.
I'm not overly optimistic given that the biggest barrier to supporting Linux has always been how much variance there is in terms of what's out there, but it's still a good thing for Linux.
In terms of perceptions of desktop Linux I don't really think it matters. Linux isn't going anywhere and as software probably has more penetration right now than any other operating system ever has.
There are many kind of linux users, and many reason to use linux. We don't all agree on the why and the how, which is one of the reason there are so many linux based OSes/distros.
And that is fine.
Yes you can on the android emulator. The biggest issue is compu arch in that case.
> I cannot compile a program on one version of the same distro and run it on another version of the same distro
Yes you can for the most part (unless it uses a capability provided by a newer kernel which is super rare and mostly limited to system tools, and less for "apps"). Actually that is what makes containers and flatpak possible. It even works accross different distros as long as cpu arch is the same.
Sure, I can see that. My solution to that one day will hopefully be to make sure devs can keep their linux platforms updated, not give up and go around it with a windows build.
But Proton discourages that, not encourages. As you said, Steam doesn't want you to know what build you are playing, and if the audience doesn't know, then the devs won't care either.
sure, Microsoft used a compatability layer to translate microsoft's API to an API Microsoft also maintains to an open source API. As far as I see it, it doesn't introduce any further points of failure that I don't already have by relying on Microsoft.
an open source compatibility layer relying on Microsoft's API... It unfortunately isn't a communicative property here.
>SteamOS solved the chicken-and-egg problem and is demonstrating that Linux-based gaming is viable.
Sure, just not in a way I feel is productive for the long term. But again, to each their own.
>2% isn't a massive number, but it might just be enough for a game developer to justify compiling a Linux-native version too.
In my mind, it reduces the need because why not just rely on Valve to do the hard "porting" work for you? It's a win-win for a dev who simply wants to launch a game. I hope your vision is the correct one, but I'm not so optimistic.
Not a personal win for someone who wants less leverage from large corporations.
[1] https://help.steampowered.com/en/faqs/view/671A-4453-E8D2-32...
[2] https://www.reddit.com/r/SteamDeck/comments/12sgnw7/what_do_...
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.
But it's simply not relevant. Linux is a success in its own right. So much better than Windows on a server. And the desktop is losing relevance (but it might bounce back when you least expect it).
The weird thing is, for me personally, is that of all the games I collected over there years when I did Windows gaming, many of them are now Apple Silicon native. Sure, the big ones like Elder Scrolls and Fallout franchises aren't there, but many others are. Even the brand-new Baldur's Gate 3 is on Apple Silicon. No Man's Sky is now too.
I'm not a big gamer anymore, but it's just interesting to see.
You can! You will need to have a UI wrapper around it on Android but if you compile it in the right way there’s no reason this can’t work (assuming they use the same CPU architecture of course)
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?
Git on Windows is only supported by installing a whole suite of Unix tools and a shell.
Tools like ccache/sccache treat windows (well msvc) as a second class citizen.
Go, the poster child for cross-compilation shatters that illusion when you need to use CGO.
Python, I believe things have gotten better but the last time I tried getting tensorflow up and running on Windows it was a long and painful path involving third party python distributions, native toolchains and changing drivers.
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.
Nope. The biggest barrier is the FUD around there being so much variance. 99% of desktop Linux is glibc-based. Beyond that, binary compatibility is no harder than Windows. Differrent yes, meaning devs used to Windows have some learning to do, but not drastically different even.
If I have to go and install one, I can do that on Windows.
You could probably get away without the mouse if you're ok with touchpad usage. But YMMV.
I've only run SteamOS on my Deck, but afaik if you install it on something else then you can still flip it between the gaming mode and the desktop mode. In desktop mode it behaves exactly like you'd expect an Arch Linux install to behave, and you can mess around with it as much as you'd like. In gaming mode it's like a console and really only plays games (but isn't limited to Steam games - people have got it running emulators and all sorts of other stuff too).
That said, I think most of us, except for the die hard purists, are fine with Proton compatibility being the main target for companies. As long as a game runs as well as it does elsewhere without restrictions or inconvenience, most of us are happy and don't care about the technical details of how it's running.
Having said this, the local library from my German city does run on GNU/Linux.
Not at all.
Docker on Windows is a shim for the Windows Jobs API, as Microsoft decided to offer the same experience instead of coming up with their own set of tooling.
In more recent Windows versions, there are other ways to manage containers, specially after containerd support improved.
The best way to distribute builds on Windows is via incredible and their VS integration.
Cross compilation never really quite works out, unless one can have a complete set of libraries and toolset of the host OS, otherwise there will always be corner cases.
Python has been quite alright when using distributions like ActiveState Python.
Git, well one cannot expect better from a SCM designed for the Linux kernel project in first place.
This is in reference to the "Year of the Linux Desktop" meme. I don't consider Steam Deck to be a general-purpose desktop computer.
Excel was acknowledged as a better software pretty early on, and could always import Lotus spreadsheets, but it really started gaining adoption once it was able to export to Lotus files.
Counter-intuitively, the ability to stay within the Lotus ecosystem (and switch back at any time) is what enabled a lot of people on the fence to try Excel out. Only once Excel became dominant did people actually switch to XLS files.
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.
Stop imagining. I'm quite literally mooching around in bed posting this from Firefox on a Steam Deck in desktop mode docked to my bedroom TV. This[1] isn't your father's Oldsmobile.
[1] https://store.steampowered.com/steamdeck/#SaleSection_24468
The pros of deciding your own runtime environment allows you to customize the system more and even run Linux on machine that has very strict resource limit.
The cons is that it is almost impossible to run a software everywhere without bundle literally anything you use into own binary. The steam itself do it(steam runtime), but I don't know if it is even close to a complete resolution because it don't really solve the problem for softwares outside of steam.
If we talk ABI's not API's, it's not a joke https://blog.hiler.eu/win32-the-only-stable-abi/
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.
[1] https://www.digitaltrends.com/computing/how-to-use-desktop-m...
It's not "vanilla", but so isn't Ubuntu, Linux Mint, Fedora Sericea, or whatever else. The underlying stack is the same, there's just customizations on top.
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.
If FreeBSD got substantial market share that would be a victory. ChromeOS gaining market share, and especially being used in education, is a defeat.
It's likely the only thing everyone can agree is Linux is whatever Linus has on his machine.
Or is it? Whether or not we have come close to TYOLOTD depends on how you define "The Desktop". If "The Desktop" is a metaphor for "the hands of the non-specialist user who doesn't even call their device 'a computer" then yeah, it's doing great. The Deck is a major step forwards in that, it wraps up Steam's entire catalog of Windows games into a nice little bubble of virtualization that varies in seamlessness depending on how new the game is.
If you define it as a desktop computer, it's nowhere near that. Most people who sit down at a a desk with a large screen to do work that involves multiple windows and a mouse pointer are still sitting in front of a box that's running Windows or MacOS. Those two operating systems still dominate that domain, as well as the world of people popping open a laptop to do windows-and-pointer work away from their office.
Linux keeps nibbling at that domain, and the work Valve did on their fork of WINE to get fifty bazillion Windows games running on this handheld Linux device is probably going to help take some bigger bites. But what do people go buy when they want A Computer? When you answer that with "Linux", that's The Year Of Linux On The Desktop.
SteamOS is actually Archlinux which is a mere Linux. Archlinux can be considered as clean sheet Linux, they avoid patching and modifications if possible. So probably even the FSF people would accept it as GNU/Linux?
Even more. The SteamDeck is using a X86_64 CPU and RDNA2 GPU. No hidden firmware locks. So it is an IBM-PC. It counts therefore 100% as Linux. I would even say 120% because it comes pre-installed.
It's funny because it's true. Valve took advantage of Microsoft API stability guarantees and executed with an overnight success 10 years in the making.
It's actually a great thing, too. You build a game once and it's more stable than any distro packaging could ever make it be.
SteamOS on the other hand is an actual GNU/Linux distribution that can be installed in any machine and any improvements that Steam does, e.g. to 3d drivers or hardware support, will benefit the entire desktop Linux ecosystem.
https://help.steampowered.com/en/faqs/view/671A-4453-E8D2-32...
It shouldn't be. For the same reasons you would shorten TCP/IP to TCP, you should shorten GNU/Linux to GNU. This avoids all of the confusion about what is "really Linux", because when people say that they actually mean GNU.
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.
I used Excel on Windows 2 from its release but I would be an outlier.
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.
Linux is the dominant operating system in tech, moreso when you include a fuzzy, but reasonable, definition -- I like to say "Linux-ish."
Kids, literally everything you encounter that isn't Windows is "Linux-ish."
Apple has found ways to make their flavor of "Linux-ish" more exclusive and that has drawbacks. In fact, each company naturally tries to figure out how they can best make money for themselves and you can judge them how you want -- but the core of this is this thing called "free and open source software," where basically everyone agrees to (or it just makes more sense to) use this stuff that's shared and out there.
You can use this as your desktop too, if you like. Here's how.
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.
Win what? It’s wildly popular in some environments, and it’s getting more popular in others. I think it’s a waste your energy to get worked up over who’s giving Linux “proper recognition”…
I can also download VirtualBox and run all Windows programs, that would mean that all Windows apps are Linux apps?
> Yes you can for the most part
You can't statically link glibc: https://github.com/cockroachdb/cockroach/issues/3392
glibc can break stuff: https://www.gamingonlinux.com/2022/08/valve-dev-understandab...
I had binaries break because the newer version if openssl was put under a slightly different name.
I'm doing exactly that with a Steam Deck. I'll be going out of town for the next few week and instead of carrying a laptop to WFH, I'm just taking my SD and a keyboard with me.
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.
That is, this is likely easily solvable, but it is most easily solvable at the beginning of a project by choice of base libraries. I can understand not wanting to change things after the fact for a presumably small user base.
If Linux is only popular on Android, Chromebooks, SteamOS, servers, embedded devices it strongly suggests to me that people are perfectly content with using Linux as long as they don't have to use the Linux desktop. (I know it's available on the Deck, but it's not the primary UI for it).
The Steam Linux numbers used to be indicative of Linux desktop popularity. If their large rise now is in large part due to the Steam Deck, that doesn't quite mean that the Linux desktop is getting popular. So it seems this is a totally fair nuance to me?
There seems to be a rising sentiment of discontent with Windows 11 and the general direction Windows is heading. There's been positive coverage of SteamDeck and the new SteamOS that is driving previously skeptical people towards being open to trying out Linux for gaming and as a general desktop OS.
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.
Sure, some might dip into desktop mode to do a bit of browsing, or set up emulators, but nobody is saying "hey, I'll just use this as my daily driver now".
Seriously, the hubris[1][2][3][4][5] of it all.
[1] https://old.reddit.com/r/SteamDeck/comments/12sgnw7/what_do_...
[2] https://old.reddit.com/r/SteamDeck/comments/12sgnw7/what_do_...
[3] https://old.reddit.com/r/SteamDeck/comments/12sgnw7/what_do_...
[4] https://old.reddit.com/r/SteamDeck/comments/12sgnw7/what_do_...
[5] https://old.reddit.com/r/SteamDeck/comments/12sgnw7/what_do_...
What if he has more than one computer and doesn't have the same setup on all of them?
Frankly I just don't see it as an interesting metric. "More Steam games are played on Steam Deck (+ desktop Linux) than on macOS." So what? Even though the Steam Deck blurs the lines a bit, I don't consider desktop gaming and console/handheld gaming to be directly comparable. Or at least, I don't consider comparing them in this way to be interesting.
Just as I don't think "Linux is the most used OS kernel on the planet" to be all that interesting. Ok, it's interesting in a certain, narrow way. Obviously its use in mobile/embedded devices and servers dwarfs its use on the desktop. If we want to compare "OS choice on servers" or "OS choice on mobile devices", that's notable.
But when we're talking about gaming, and we want to compare "desktop macOS gamers" to other things, the only other things I think are relevant are "desktop Windows gamers" and "desktop Linux gamers".
If Steam-using desktop Linux gamers were surpassing Steam-using desktop macOS gamers, that would be huge, important news! But a bunch of people buying a Steam Deck and raising the number of people gaming on the Linux kernel is, to me, no more interesting than saying there are a ton of people playing games on the Linux kernel because there are a ton of people who game on Android phones. Whoop-de-do, film at 11.
(Full disclosure: I'm a desktop Linux user, and have been for over 20 years.)
They claimed that they'd need a different implementation per distribution. Which makes no sense. It's just open()/ioctl_tty()/read()/write(), all of which are in the libc of every distribution that has ever existed since the 90s.
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
There was a working, officially supported Miracast app on the Xbox One for a while and I used that for a screen extender of a Windows 10 Phone and a laptop for a few years before the app stopped working. I think it stopped working only because no one was using it.
The entire Ruby ecosystem has never had great Windows DX: the most recommended Windows builds of Ruby itself come from a "second-party" group because official upstream doesn't seem to care to bundle nice installers, there's a ton of papercuts on Windows in the base libraries even in APIs you don't think would be platform-dependent, and a large number of third-party libraries and apps just kind of invariantly assume that you will never try to use them on Windows.
WINE is not even an emulator, or so its name says.
The only reason why SteamOS might conceivably not count, is because the users didn't explicitly, knowingly choose Linux. They chose Steam. Valve pushes the Steam deck to its users, and those users buy it, possibly not even caring that it's running Linux rather than Windows.
Doesn't matter though. It's still Linux, even if the users may not be aware of it.
At least that's my very limited experience with linux gaming. I started only a couple of weeks ago, but so far, everything works incredibly well.
So less avoiding Windows licences and more avoinding getting cut out by MS.
There are a couple of other game launchers, like Heroic, which handles both Epic and GOG games, which also works great.
But what's even better, is that if a GOG game doesn't quite run perfectly on Heroic because it lacks some of the latest refinements from Proton, you can import that game into Steam and taie advantage of Valve's nice work on Proton anyway.
I might do that; Cyberpunk on Heroic doesn't support raytracing or DLSS, and as I understand it, those do work if I were to run it on Steam. So that's a very nice option to have.
That won't be for a while, though. But im the mean time, Wine/Proton/DXVK means that the Linux experience is excellent even without explicit developer support, so the Linux user base can grow beyond just the handful of die hards. Anyone can easily game on Linux now.
That is how Steam solves the chicken/egg problem.
You can. Idr if Docker Desktop supports it or not, but you can install Docker Engine for Windows and plug it into the Docker CLI and all that for sure.
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.
In the mid '10s Feral (and a to a lesser extent Aspyr), ported many AAA games to linx. Many of the ports were of quite good quality. I own a few of them, yet occasionally I have to switch to the Proton version as the native one fails to start.
OSS games are the exception of course: being able to produce a good working binary from source make them future proof.
My biggest issue with Windows was the forced reboots.
There are reasons to want to talk about Linux, but it probably shouldn't be much more common than NT in day to day vocabulary.
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.