No, but I don't think the FOSS folk will like the two paths beyond the duopoly. Path 1: China. China uses a lot of Windows, and China does not like that dependence on a US company providing what...
No, but I don't think the FOSS folk will like the two paths beyond the duopoly.
Path 1: China. China uses a lot of Windows, and China does not like that dependence on a US company providing what amounts to critical infrastructure. Deepin is one example of this push. I think China is going to pour billions of dollars into making a competitive consumer OS, and pour billions into writing enough user space apps that people use it. And the Chinese government has both the money and the "fuck you do what we want" ability to force people in China to switch.
Path 2: Web. Internet browsers have trended into becoming their own virtualized, mini OS's. And more and more people are using almost all web apps, like google docs, and the rest of the g-suite, for their computing needs. And to some extent I appreciate this; Spotify works on Linux despite it being maintained by literal volunteers at the company just because it's fundamentally a react app that works everywhere chromium works.
So 1 involves an OS owned by the Chinese government, and 2 involves basically an OS dictated by Google.
I wonder how long off that's going to be given the many pirate copies of WinXP running in China everywhere (local cops, universities, etc) . I guess for the majority of Chinese , as long as...
I wonder how long off that's going to be given the many pirate copies of WinXP running in China everywhere (local cops, universities, etc) . I guess for the majority of Chinese , as long as there's a webbrowser , QQ , some Office-esque software (WPS), they'd probably be happy enough . There's still a fairly large contingent of netbar attendees who play games on Windows though .
MS and Apple arguably don't own personal computing already. There are more Android devices than Win/MacOS combined. Chromebooks are a thing as well. I know loads of people who don't even own...
MS and Apple arguably don't own personal computing already. There are more Android devices than Win/MacOS combined. Chromebooks are a thing as well. I know loads of people who don't even own desktop/laptop machines any more, just a phone and a tablet.
Even as an avid user of such for over 20 years now, I don't see Linux on the desktop being much more of a thing (or less of a thing) because it doesn't solve a problem most people have. Rather, it creates a problem for them, the resolution of which leaves them almost exactly where they were in the first place. Eg: the problem is installing an OS, which is a huge hurdle for a lot of people no matter how easy we know it is to run the Debian/ubuntu/whatever install - just to get a web browser, which windows already offers them.
With the possible exception of the developing world but I don't know how things are going, I suspect they'll bypass desktop computing entirely for the most part, going straight to tablets.
The question I'm interested in is does this matter? Does having a few organisations doing most of a thing matter? I don't see that it does, particularly. I mean sure, Windows is a nightmare, but it does mostly work. Mostly. Enough for most people. Apple are price gouging motherfuckers but their stuff works, and works is what matters.
I don't think they are. Their computers work. They don't work as well as our computers do but does that matter? Stuff still gets done. Could MS/Apple/etc be better? Sure they could. I'm not saying...
I don't think they are. Their computers work. They don't work as well as our computers do but does that matter? Stuff still gets done.
Could MS/Apple/etc be better? Sure they could. I'm not saying they're ideal because they're not. But they are perhaps good enough.
Three out of four. But I'm assuming you have a few android devices here and there too. Me neither. But I'm not worried that's going to happen. Linux as a server/embedded/etc OS isn't going...
all three OSes
Three out of four. But I'm assuming you have a few android devices here and there too.
I really, really don't want to live in a world where it's required for me to use Windows.
Me neither. But I'm not worried that's going to happen. Linux as a server/embedded/etc OS isn't going anywhere, and there'll always been enough motivated hackers to keep the desktop projects alive. I think I read somewhere that there are more linux desktop users than there ever have been before, but proportionally that amount is about the same as it's always been. I could be wrong about that, I'll try to find the article later when I'm back at my linux machine..
There's only a very blurry line between desktop and handheld for a lot of home users, if any line at all. I don't think for personal home computing it's a particularly meaningful distinction any...
There's only a very blurry line between desktop and handheld for a lot of home users, if any line at all. I don't think for personal home computing it's a particularly meaningful distinction any more.
Also don't forget Android is open source so far from being a lost cause, I'd say that's a massive win.
Android complies with the relevant open source licenses (GPL and Apache, I think), that's all it needs to do to be open source. But regardless of that it's extremely customisable - thousands of...
Android complies with the relevant open source licenses (GPL and Apache, I think), that's all it needs to do to be open source.
But regardless of that it's extremely customisable - thousands of different sorts of device run Android in countless different configurations - although being customisable is not a requirement to be FOSS (have you used Gnome Shell recently?); it's reasonably secure when deployed properly (heck, Windows is secure when set up right), although equally, not a FOSS requirement; and it's as private as you want it to be. You don't have to install the Google/Samsung/whatever overlays which I suspect you believe to violate your privacy.
I don't disagree that privacy and security are nice to have and I'd like to see them in all software regardless of license. Customisability I'm less sold on, at least when it comes to end user interface. Sometimes giving people options makes things harder to use for little benefit. But that's a rather different discussion.
However, I think it's a bit unreasonable to say Android doesn't have those things. Even Windows has those things and so does MacOS, albeit to differing degrees.
Open source gets you open source. That's all. You might want it to get you other things but there's no promises of those any more than you'd get with closed source software. The reason I mentioned...
Open source gets you open source. That's all. You might want it to get you other things but there's no promises of those any more than you'd get with closed source software. The reason I mentioned Shell was the devs keep taking stuff out of it (and gnome in general). Every release seems to remove or hide away more options, takes out features - which is all fine from a usability perspective but people often seem to get cross about it.
I've been using Shell since it hit Debian's testing repo. Almost all the extensions I've tried are unreliable and invariably break every time Gnome bumps it's version number. Apparently writing them isn't as easy as you think... :) (also, yes, why on earth did they choose JS? Gnome devs made weird choices) Eventually I gave up with most non-stock stuff but I'm not really bothered because Shell is the most invisible DE I've ever used.
To be honest, being able to change a few UI elements is not what I thought you meant by customisable, I thought you were talking about changing things at a more fundamental level - the kind of customising you'd do to Android to make it run on your TV rather than on your phone. I suppose that would probably be better expressed as being portable rather than customisable, but anyway, sorry for misunderstanding.
On the customisability front, I would argue that all the OSes are pretty customisable at the end user level. You can get all sort of UI tweaks up to and including full shell replacements for Windows, I don't know about MacOS these days but back in my OSX days I had some fairly heavy customisation going on (because I had a lot of time bored at work to fiddle about with it..)
We might never hit "Year of the Linux Desktop," but I think one day we'll wake up and see we're there, if only because the vast majority of the masses have ditched proper general-purpose computers...
We might never hit "Year of the Linux Desktop," but I think one day we'll wake up and see we're there, if only because the vast majority of the masses have ditched proper general-purpose computers in favor of locked-down media consumption devices.
Linux on the desktop does not see the kinds of back-slides that especially plagues Windows. The programs progressively get better, and the kind of performance bloat is much slower, allowing hardware to last longer. Each year, it seems more and more resources are being allocated to enable Linux to explode in popularity real quick. Right now, if you're a power user (have ever opened the registry editor on Windows), there's virtually nothing stopping you from using Linux on the desktop.
AMDGPU flipped Linux support for GPU on its head. Now that AMD is putting out genuinely competitive GPU again, this really opens up some interesting possiblities for Linux on the Desktop. Especially since the PS5 is basically Zen3 + RX 6000 series.
I see Linux going places. I foresee Valve building on SteamOS to build a VR OS to free them from Microsoft's thumb. They do that, and Linux could explode on desktop, if only because once you have Linux and a C compiler, all the rest of the ecosystem can follow. Presuming non-locked bootloaders and sufficient disk space of course.
There is literally nothing with WSL, you don't even have to lose your Windows. The year of the Linux Desktop happens when there is a Windows 10 update that runs primarily off a Linux Kernel, uses...
Right now, if you're a power user (have ever opened the registry editor on Windows), there's virtually nothing stopping you from using Linux on the desktop.
There is literally nothing with WSL, you don't even have to lose your Windows.
The year of the Linux Desktop happens when there is a Windows 10 update that runs primarily off a Linux Kernel, uses virtualization and Microsoft-amped mega Wine for everything else, and suddenly the latest update for the most popular consumer operating system is based on Linux. That day is coming, and it will probably be considered an abomination in FOSS circles, but that's what it would take.
WSL is only available to Pro or higher. Losing Windows is a feature. Especially if you care one iota about privacy. I think Valve is really the key, as it boils down to GUI and OEM adoption. If...
I think Valve is really the key, as it boils down to GUI and OEM adoption. If Valve takes another stab at a SteamOS with a VR focus at the right time, making the UI open the same way they did Proton, it could blow the doors wide open for adoption, esp if offering incentives for OEM builders. We'd see a systemd-style migration of popular distros to using the Valve-written GUI in a heartbeat. And people with existing hardware that wanted that nice VR UI might go through the 4 steps required to install it (download, double click installer, follow prompt, reboot)
Companies might live longer than people, but they don't live nearly as long as species. While we're still dealing with their colonial legacies, the various East India companies no longer control...
Companies might live longer than people, but they don't live nearly as long as species. While we're still dealing with their colonial legacies, the various East India companies no longer control any territory. Bell Telephone no longer owns the phone network, and nobody is still worried about the power of Standard Oil.
Capitalism might manage to guarantee us a certain level of monopilization and badness, but, as you note, Chromebooks are threatening Microsoft and Apple. Google is getting investigated over antitrust violations, which could change who controls Chromebooks. Apple is trying to move their whole system to Arm, which could blow up in their face. Microsoft is pushing Windows as a Microsoft Account terminal, which seems to be succeeding but which no one seems to actually like. Phones and VR and "computers you yell at" (AKA Alexa) are waiting in the wings to possibly eat everybody's lunch if they manage to get good enough at traditional desktop tasks. We haven't actually reached a steady state.
FWIW at my job the duopoly is between Mac and Ubuntu. The Macs have fewer weird hardware problems, but more weird software problems. The Linux machines build all our software no problem and hardly ever complain that they can't find printf or need a 40 GB IDE installed, but when new people install Linux on their machines to get work done they run into new and unique driver issues. Still sort of a uniform level of broken with two viable options, but none of it is owned by Microsoft.
Just a note, but that's pretty unlikely. There's a reason why there are literally 0 AGs from blue states on the case: all of them considered the case brought by the federal AG to be awful, rushed,...
Google is getting investigated over antitrust violations, which could change who controls Chromebooks.
Just a note, but that's pretty unlikely. There's a reason why there are literally 0 AGs from blue states on the case: all of them considered the case brought by the federal AG to be awful, rushed, and unlikely to succeed, likely pushed so that they could announce it before the election.
It doesn't mention most if not all of Google's actual anticompetitive behavior.
Google has entered the chat I think the next alternative is our current major, hardware-dedicated operating system. The dominant one we have is Android/Chrome (now essentially the same ecosystem),...
Google has entered the chat
I think the next alternative is our current major, hardware-dedicated operating system. The dominant one we have is Android/Chrome (now essentially the same ecosystem), and people generally use tablets and phones more than desktop computers, and dedicated Android/Chrome hardware for a lot of tasks.
I would love to see a major GNU/Linux player, but I also feel the major distros we have are not good for the market. Ubuntu keeps making iffy calls about their technology every couple of years(Unity! Not Unity, GNOME! Snap, to replace Debian-based packages!), Fedora is intended as a stable testbed, and SUSE just doesn't have the Linux community market share. The distros I think make the best desktop OSes are either community distros (Mint, MX) or super niche (Arch). Mint may be the best player, but they're dependent on Ubuntu, so automatically slightly fragile.
If we could get a Mint type distro built on an upstream distro like Debian (we have LMDE, but it's more of a testbed than a major distro), we'd be headed in the right direction. Something that is ideologically consumer-tier with an LTS release cycle would be a hard requirment, with essential apps (desktop, browsers, office suite) updated regularly, or even just somehow getting Office to Linux. From there, we could probably just encourage the use of AppImage for OSX-style application install/use, as well as distro-neutral stuff like flatpak as an option (but, we need to fix GUI integration for flatpak).
A favorite quote (or rather, paraphrase of mine) is from Linus Torvalds, who basically said people don't use operating systems, they use applications. I would prefer the operating system layer these apps run on be an open one, but most people don't care until they actively seek to make a conscious effort. I guess my point is for the majority of users, it's about what they can do on their computer, rather than what they can do in their operating system.
Personal computing for the masses as we know it, I think yes we are stuck with the duopoly forever. Even with paradigm changes, such as cloud computing where our devices are basically terminals to...
Personal computing for the masses as we know it, I think yes we are stuck with the duopoly forever. Even with paradigm changes, such as cloud computing where our devices are basically terminals to someone's mainframe, people still need a device with a OS for that and the majority of users will use macOS or Windows. The closest we managed to get away from the duopoly is Android, I guess.
Even developers, the ones we usually think of as power users, will be using the browser to program via Github and deploy code in the next years. The convenience will be so great that most of us wont bother to configure their own machine/OS.
I also have a theory, that some years from now (decades to be more precise) this companies are going to grow so large, that consumers are going to buy "Apple houses" or "Amazon houses", where everything inside is highly integrated with those companies' proprietary networks. Their interest for automobiles was the first step.
Docker made Linux even more popular on the server. I speculate that something like that might happen on the client where there is a cross-platform packaging format for apps that run in a VM. You...
Docker made Linux even more popular on the server. I speculate that something like that might happen on the client where there is a cross-platform packaging format for apps that run in a VM.
You can see the beginnings of this with Linux-within-VM support on Chrome OS and Windows that should soon be good enough for many kinds of software development. Also, I think game emulators work this way? I’m most familiar with how interactive fiction works, where files uploaded to if-archive never change and it’s up to emulators to support them. But that’s an easy case because it was a well-defined format and virtual machine from the beginning.
The web sort of does this but it’s a very complicated API that’s constantly changing, and apps aren’t self-contained. Web Bundles might do it, but ugh. Android apks have a similar ugh factor. WebAssembly seems like a nice foundation but lacks api’s for UI stuff unless you count web api’s. I have some hope that something nice might evolve out of Flutter.
If such a format became popular, which OS you run would become much less important.
The reason this hasn’t happened yet is that modern UI’s are incredibly complicated particularly due to internationalization and accessibility. Some is accidental complexity but much of the complexity is embedded in human culture. Incomplete solutions don’t have the traction.
Valve's Steam Proton setup (basically their own WINE distribution) looks to be shaping up into an offensively ugly but in practice quite serviceable game platform. Write once for the...
Valve's Steam Proton setup (basically their own WINE distribution) looks to be shaping up into an offensively ugly but in practice quite serviceable game platform. Write once for the well-understood subset of Windows/DirectX that Proton can actually support, and run anywhere.
I don't think the DMCA figures into it much currently, although the Oracle/Google case on API copyrightability and whether you can use copyrighted material as necessary for compatibility might affect things.
I would say "No but also yes?" I think its going to be a triopoly between Mac, Windows, and ChromeBook. I think the only reason its not going to stay a duopoly is only because Google (and possibly...
I would say "No but also yes?" I think its going to be a triopoly between Mac, Windows, and ChromeBook. I think the only reason its not going to stay a duopoly is only because Google (and possibly Amazon) have such obscene amounts of "fuck you" money.
Let's start off by saying: Microsoft and Apple at this point have massive advantages. Their OS's are household names. They ran entire ad campaigns about how they are the only OS's, despite Linux existing. If you go to a Best Buy or Walmart or whatever retail store that sells laptops, your options are Mac, Windows, or Chromebook. The amount of money those three companies can blow on hardware, OS improvements, marketing, etc might as well be limitless. They will be limited by how successful/necessary they are, but if one of those companies decided they wanted to, they could throw millions to billions into it.
So from that standpoint, it is already an uphill battle. I'm going to bullet point the rest of my arguments because I really need to get back to work but some other things:
They would need to start having laptops/desktops with Linux pre-installed available in all major retailers. People aren't going to want to install their own OS. We can argue until the cows come home about how easy or hard it is, but if it was a matter of difficulty Linux would be more common by now.
"I installed Linux on my parents computer and its going fine." Ok yeah, but what happens when this laptop dies? Are you going to go online shopping to find a new laptop for your family members that meets their needs in their price range, buy it, install Linux on it, move all their old files to the new laptop, make sure the laptop works as expected with the hardware they use, and give it back to them? Repeat for every person you show Linux to? I guess some people will, but not enough for Linux to get traction to go mainstream. Maybe the argument is once you get them on Linux once they'll see its not so bad and will try to install the OS themselves next time they need a new laptop? Another thing I highly doubt.
There would need to be an active marketing campaign that explains what Linux is and why they should buy a Linux laptop (operating under the assumption that my first point happens). People use chromebooks because they're cheap, they meet the needs of their target users, and Google is a company name that they recognize and trust in the tech space. Most people have never heard of Linux, and even if they have a lot haven't used it or would trust Linux as a company over Windows and Mac.
FOSS isn't a compelling argument to lots of people. If it was, people would have gotten off of Facebook and Instagram. People would be de-googling in bulk. I'd be able to get my friends and family to use signal. People don't care or understand why open-source is better than close-sourced. In fact, I've had to explain to people multiple times why open-source can make your software more secure and they almost never buy it.
The market of people who want a Linux laptop and would rather buy it over build it is small. I'm in that target demographic, but I don't really know anyone else who is the same way as me.
I probably have more reasons that I'll come back and comment after I have more time to reflect and think. That's what I've got off-the-cuff tho.
I'll say that I don't really think FOSS is the only competitor to those 3. I mention FOSS because it has inevitably come up in every discussion I've had with tech people about why Linux...
I'll say that I don't really think FOSS is the only competitor to those 3. I mention FOSS because it has inevitably come up in every discussion I've had with tech people about why Linux will/wont/should/etc become mainstream. Knowing the demographics breakdown of Tildes, especially in careers, I just figured I'd get out ahead of it.
Really the only companies I can see competing in this space are just as bad or worse version of what we already have. Facebook, Amazon, and Tencent/China are really the only ones that come to mind for me that have the same capital to throw at the problem as Apple, Microsoft, and Google.
I guess there is a 3rd alternative, which is the/a government decided to start going anti-trust on a bunch of tech companies asses, leveling the playing field for other companies to only have to overcome inertia and name recognition. Which is still a big task but at least its not having to do that while at a massive money disadvantage.
Depending on your use case, the fact that only a small % of computer users run Linux does not affect you much. I use Debian daily. It's not perfect but if you really don't want to use Windows or...
Every time a desktop operating system other than Windows or Mac OS is mentioned - Linux, BSD, Haiku, etc - the prevailing opinion seems to range from "this is a waste of time" to "this is interesting but not useful".
Depending on your use case, the fact that only a small % of computer users run Linux does not affect you much. I use Debian daily. It's not perfect but if you really don't want to use Windows or OSX it's not too painful. Granted, I mainly just browse the web, watch videos and write code on my personal computer.
I'd want to challenge the implicit assumption that the desktop will maintain relevance in personal computing. For work, yes, people will need company-provided laptops or desktops. But for personal...
I'd want to challenge the implicit assumption that the desktop will maintain relevance in personal computing.
For work, yes, people will need company-provided laptops or desktops. But for personal computing? If someone set their mind to it, a phone and a smart TV would be enough. If you don't game (on a PC) or don't have a hobby that requires a computer, you really don't need a desktop at all.
The duopoly will last only as long as the current paradigm. If desktop computing becomes more niche, I expect that its more technical userbase will be more likely to use linux.
No, but I don't think the FOSS folk will like the two paths beyond the duopoly.
Path 1: China. China uses a lot of Windows, and China does not like that dependence on a US company providing what amounts to critical infrastructure. Deepin is one example of this push. I think China is going to pour billions of dollars into making a competitive consumer OS, and pour billions into writing enough user space apps that people use it. And the Chinese government has both the money and the "fuck you do what we want" ability to force people in China to switch.
Path 2: Web. Internet browsers have trended into becoming their own virtualized, mini OS's. And more and more people are using almost all web apps, like google docs, and the rest of the g-suite, for their computing needs. And to some extent I appreciate this; Spotify works on Linux despite it being maintained by literal volunteers at the company just because it's fundamentally a react app that works everywhere chromium works.
So 1 involves an OS owned by the Chinese government, and 2 involves basically an OS dictated by Google.
That's worse than what we have now, tbh.
I wonder how long off that's going to be given the many pirate copies of WinXP running in China everywhere (local cops, universities, etc) . I guess for the majority of Chinese , as long as there's a webbrowser , QQ , some Office-esque software (WPS), they'd probably be happy enough . There's still a fairly large contingent of netbar attendees who play games on Windows though .
MS and Apple arguably don't own personal computing already. There are more Android devices than Win/MacOS combined. Chromebooks are a thing as well. I know loads of people who don't even own desktop/laptop machines any more, just a phone and a tablet.
Even as an avid user of such for over 20 years now, I don't see Linux on the desktop being much more of a thing (or less of a thing) because it doesn't solve a problem most people have. Rather, it creates a problem for them, the resolution of which leaves them almost exactly where they were in the first place. Eg: the problem is installing an OS, which is a huge hurdle for a lot of people no matter how easy we know it is to run the Debian/ubuntu/whatever install - just to get a web browser, which windows already offers them.
With the possible exception of the developing world but I don't know how things are going, I suspect they'll bypass desktop computing entirely for the most part, going straight to tablets.
The question I'm interested in is does this matter? Does having a few organisations doing most of a thing matter? I don't see that it does, particularly. I mean sure, Windows is a nightmare, but it does mostly work. Mostly. Enough for most people. Apple are price gouging motherfuckers but their stuff works, and works is what matters.
I don't think they are. Their computers work. They don't work as well as our computers do but does that matter? Stuff still gets done.
Could MS/Apple/etc be better? Sure they could. I'm not saying they're ideal because they're not. But they are perhaps good enough.
Three out of four. But I'm assuming you have a few android devices here and there too.
Me neither. But I'm not worried that's going to happen. Linux as a server/embedded/etc OS isn't going anywhere, and there'll always been enough motivated hackers to keep the desktop projects alive. I think I read somewhere that there are more linux desktop users than there ever have been before, but proportionally that amount is about the same as it's always been. I could be wrong about that, I'll try to find the article later when I'm back at my linux machine..
There's only a very blurry line between desktop and handheld for a lot of home users, if any line at all. I don't think for personal home computing it's a particularly meaningful distinction any more.
Also don't forget Android is open source so far from being a lost cause, I'd say that's a massive win.
Android complies with the relevant open source licenses (GPL and Apache, I think), that's all it needs to do to be open source.
But regardless of that it's extremely customisable - thousands of different sorts of device run Android in countless different configurations - although being customisable is not a requirement to be FOSS (have you used Gnome Shell recently?); it's reasonably secure when deployed properly (heck, Windows is secure when set up right), although equally, not a FOSS requirement; and it's as private as you want it to be. You don't have to install the Google/Samsung/whatever overlays which I suspect you believe to violate your privacy.
I don't disagree that privacy and security are nice to have and I'd like to see them in all software regardless of license. Customisability I'm less sold on, at least when it comes to end user interface. Sometimes giving people options makes things harder to use for little benefit. But that's a rather different discussion.
However, I think it's a bit unreasonable to say Android doesn't have those things. Even Windows has those things and so does MacOS, albeit to differing degrees.
Open source gets you open source. That's all. You might want it to get you other things but there's no promises of those any more than you'd get with closed source software. The reason I mentioned Shell was the devs keep taking stuff out of it (and gnome in general). Every release seems to remove or hide away more options, takes out features - which is all fine from a usability perspective but people often seem to get cross about it.
I've been using Shell since it hit Debian's testing repo. Almost all the extensions I've tried are unreliable and invariably break every time Gnome bumps it's version number. Apparently writing them isn't as easy as you think... :) (also, yes, why on earth did they choose JS? Gnome devs made weird choices) Eventually I gave up with most non-stock stuff but I'm not really bothered because Shell is the most invisible DE I've ever used.
To be honest, being able to change a few UI elements is not what I thought you meant by customisable, I thought you were talking about changing things at a more fundamental level - the kind of customising you'd do to Android to make it run on your TV rather than on your phone. I suppose that would probably be better expressed as being portable rather than customisable, but anyway, sorry for misunderstanding.
On the customisability front, I would argue that all the OSes are pretty customisable at the end user level. You can get all sort of UI tweaks up to and including full shell replacements for Windows, I don't know about MacOS these days but back in my OSX days I had some fairly heavy customisation going on (because I had a lot of time bored at work to fiddle about with it..)
We might never hit "Year of the Linux Desktop," but I think one day we'll wake up and see we're there, if only because the vast majority of the masses have ditched proper general-purpose computers in favor of locked-down media consumption devices.
Linux on the desktop does not see the kinds of back-slides that especially plagues Windows. The programs progressively get better, and the kind of performance bloat is much slower, allowing hardware to last longer. Each year, it seems more and more resources are being allocated to enable Linux to explode in popularity real quick. Right now, if you're a power user (have ever opened the registry editor on Windows), there's virtually nothing stopping you from using Linux on the desktop.
Some relatively recent huge advancements:
I see Linux going places. I foresee Valve building on SteamOS to build a VR OS to free them from Microsoft's thumb. They do that, and Linux could explode on desktop, if only because once you have Linux and a C compiler, all the rest of the ecosystem can follow. Presuming non-locked bootloaders and sufficient disk space of course.
There is literally nothing with WSL, you don't even have to lose your Windows.
The year of the Linux Desktop happens when there is a Windows 10 update that runs primarily off a Linux Kernel, uses virtualization and Microsoft-amped mega Wine for everything else, and suddenly the latest update for the most popular consumer operating system is based on Linux. That day is coming, and it will probably be considered an abomination in FOSS circles, but that's what it would take.
WSL is only available to Pro or higher.
Losing Windows is a feature. Especially if you care one iota about privacy.
I think Valve is really the key, as it boils down to GUI and OEM adoption. If Valve takes another stab at a SteamOS with a VR focus at the right time, making the UI open the same way they did Proton, it could blow the doors wide open for adoption, esp if offering incentives for OEM builders. We'd see a systemd-style migration of popular distros to using the Valve-written GUI in a heartbeat. And people with existing hardware that wanted that nice VR UI might go through the 4 steps required to install it (download, double click installer, follow prompt, reboot)
Companies might live longer than people, but they don't live nearly as long as species. While we're still dealing with their colonial legacies, the various East India companies no longer control any territory. Bell Telephone no longer owns the phone network, and nobody is still worried about the power of Standard Oil.
Capitalism might manage to guarantee us a certain level of monopilization and badness, but, as you note, Chromebooks are threatening Microsoft and Apple. Google is getting investigated over antitrust violations, which could change who controls Chromebooks. Apple is trying to move their whole system to Arm, which could blow up in their face. Microsoft is pushing Windows as a Microsoft Account terminal, which seems to be succeeding but which no one seems to actually like. Phones and VR and "computers you yell at" (AKA Alexa) are waiting in the wings to possibly eat everybody's lunch if they manage to get good enough at traditional desktop tasks. We haven't actually reached a steady state.
FWIW at my job the duopoly is between Mac and Ubuntu. The Macs have fewer weird hardware problems, but more weird software problems. The Linux machines build all our software no problem and hardly ever complain that they can't find printf or need a 40 GB IDE installed, but when new people install Linux on their machines to get work done they run into new and unique driver issues. Still sort of a uniform level of broken with two viable options, but none of it is owned by Microsoft.
Just a note, but that's pretty unlikely. There's a reason why there are literally 0 AGs from blue states on the case: all of them considered the case brought by the federal AG to be awful, rushed, and unlikely to succeed, likely pushed so that they could announce it before the election.
It doesn't mention most if not all of Google's actual anticompetitive behavior.
The existence of the anticompetitive behavior supports an actual antitrust intervention, though, even if this particular investigation is bogus.
Google has entered the chat
I think the next alternative is our current major, hardware-dedicated operating system. The dominant one we have is Android/Chrome (now essentially the same ecosystem), and people generally use tablets and phones more than desktop computers, and dedicated Android/Chrome hardware for a lot of tasks.
I would love to see a major GNU/Linux player, but I also feel the major distros we have are not good for the market. Ubuntu keeps making iffy calls about their technology every couple of years(Unity! Not Unity, GNOME! Snap, to replace Debian-based packages!), Fedora is intended as a stable testbed, and SUSE just doesn't have the Linux community market share. The distros I think make the best desktop OSes are either community distros (Mint, MX) or super niche (Arch). Mint may be the best player, but they're dependent on Ubuntu, so automatically slightly fragile.
If we could get a Mint type distro built on an upstream distro like Debian (we have LMDE, but it's more of a testbed than a major distro), we'd be headed in the right direction. Something that is ideologically consumer-tier with an LTS release cycle would be a hard requirment, with essential apps (desktop, browsers, office suite) updated regularly, or even just somehow getting Office to Linux. From there, we could probably just encourage the use of AppImage for OSX-style application install/use, as well as distro-neutral stuff like flatpak as an option (but, we need to fix GUI integration for flatpak).
A favorite quote (or rather, paraphrase of mine) is from Linus Torvalds, who basically said people don't use operating systems, they use applications. I would prefer the operating system layer these apps run on be an open one, but most people don't care until they actively seek to make a conscious effort. I guess my point is for the majority of users, it's about what they can do on their computer, rather than what they can do in their operating system.
Personal computing for the masses as we know it, I think yes we are stuck with the duopoly forever. Even with paradigm changes, such as cloud computing where our devices are basically terminals to someone's mainframe, people still need a device with a OS for that and the majority of users will use macOS or Windows. The closest we managed to get away from the duopoly is Android, I guess.
Even developers, the ones we usually think of as power users, will be using the browser to program via Github and deploy code in the next years. The convenience will be so great that most of us wont bother to configure their own machine/OS.
I also have a theory, that some years from now (decades to be more precise) this companies are going to grow so large, that consumers are going to buy "Apple houses" or "Amazon houses", where everything inside is highly integrated with those companies' proprietary networks. Their interest for automobiles was the first step.
Do you find it unlikely?
Docker made Linux even more popular on the server. I speculate that something like that might happen on the client where there is a cross-platform packaging format for apps that run in a VM.
You can see the beginnings of this with Linux-within-VM support on Chrome OS and Windows that should soon be good enough for many kinds of software development. Also, I think game emulators work this way? I’m most familiar with how interactive fiction works, where files uploaded to if-archive never change and it’s up to emulators to support them. But that’s an easy case because it was a well-defined format and virtual machine from the beginning.
The web sort of does this but it’s a very complicated API that’s constantly changing, and apps aren’t self-contained. Web Bundles might do it, but ugh. Android apks have a similar ugh factor. WebAssembly seems like a nice foundation but lacks api’s for UI stuff unless you count web api’s. I have some hope that something nice might evolve out of Flutter.
If such a format became popular, which OS you run would become much less important.
The reason this hasn’t happened yet is that modern UI’s are incredibly complicated particularly due to internationalization and accessibility. Some is accidental complexity but much of the complexity is embedded in human culture. Incomplete solutions don’t have the traction.
Valve's Steam Proton setup (basically their own WINE distribution) looks to be shaping up into an offensively ugly but in practice quite serviceable game platform. Write once for the well-understood subset of Windows/DirectX that Proton can actually support, and run anywhere.
I don't think the DMCA figures into it much currently, although the Oracle/Google case on API copyrightability and whether you can use copyrighted material as necessary for compatibility might affect things.
I would say "No but also yes?" I think its going to be a triopoly between Mac, Windows, and ChromeBook. I think the only reason its not going to stay a duopoly is only because Google (and possibly Amazon) have such obscene amounts of "fuck you" money.
Let's start off by saying: Microsoft and Apple at this point have massive advantages. Their OS's are household names. They ran entire ad campaigns about how they are the only OS's, despite Linux existing. If you go to a Best Buy or Walmart or whatever retail store that sells laptops, your options are Mac, Windows, or Chromebook. The amount of money those three companies can blow on hardware, OS improvements, marketing, etc might as well be limitless. They will be limited by how successful/necessary they are, but if one of those companies decided they wanted to, they could throw millions to billions into it.
So from that standpoint, it is already an uphill battle. I'm going to bullet point the rest of my arguments because I really need to get back to work but some other things:
They would need to start having laptops/desktops with Linux pre-installed available in all major retailers. People aren't going to want to install their own OS. We can argue until the cows come home about how easy or hard it is, but if it was a matter of difficulty Linux would be more common by now.
"I installed Linux on my parents computer and its going fine." Ok yeah, but what happens when this laptop dies? Are you going to go online shopping to find a new laptop for your family members that meets their needs in their price range, buy it, install Linux on it, move all their old files to the new laptop, make sure the laptop works as expected with the hardware they use, and give it back to them? Repeat for every person you show Linux to? I guess some people will, but not enough for Linux to get traction to go mainstream. Maybe the argument is once you get them on Linux once they'll see its not so bad and will try to install the OS themselves next time they need a new laptop? Another thing I highly doubt.
There would need to be an active marketing campaign that explains what Linux is and why they should buy a Linux laptop (operating under the assumption that my first point happens). People use chromebooks because they're cheap, they meet the needs of their target users, and Google is a company name that they recognize and trust in the tech space. Most people have never heard of Linux, and even if they have a lot haven't used it or would trust Linux as a company over Windows and Mac.
FOSS isn't a compelling argument to lots of people. If it was, people would have gotten off of Facebook and Instagram. People would be de-googling in bulk. I'd be able to get my friends and family to use signal. People don't care or understand why open-source is better than close-sourced. In fact, I've had to explain to people multiple times why open-source can make your software more secure and they almost never buy it.
The market of people who want a Linux laptop and would rather buy it over build it is small. I'm in that target demographic, but I don't really know anyone else who is the same way as me.
I probably have more reasons that I'll come back and comment after I have more time to reflect and think. That's what I've got off-the-cuff tho.
I'll say that I don't really think FOSS is the only competitor to those 3. I mention FOSS because it has inevitably come up in every discussion I've had with tech people about why Linux will/wont/should/etc become mainstream. Knowing the demographics breakdown of Tildes, especially in careers, I just figured I'd get out ahead of it.
Really the only companies I can see competing in this space are just as bad or worse version of what we already have. Facebook, Amazon, and Tencent/China are really the only ones that come to mind for me that have the same capital to throw at the problem as Apple, Microsoft, and Google.
I guess there is a 3rd alternative, which is the/a government decided to start going anti-trust on a bunch of tech companies asses, leveling the playing field for other companies to only have to overcome inertia and name recognition. Which is still a big task but at least its not having to do that while at a massive money disadvantage.
Depending on your use case, the fact that only a small % of computer users run Linux does not affect you much. I use Debian daily. It's not perfect but if you really don't want to use Windows or OSX it's not too painful. Granted, I mainly just browse the web, watch videos and write code on my personal computer.
I'd want to challenge the implicit assumption that the desktop will maintain relevance in personal computing.
For work, yes, people will need company-provided laptops or desktops. But for personal computing? If someone set their mind to it, a phone and a smart TV would be enough. If you don't game (on a PC) or don't have a hobby that requires a computer, you really don't need a desktop at all.
The duopoly will last only as long as the current paradigm. If desktop computing becomes more niche, I expect that its more technical userbase will be more likely to use linux.