Look at the Steam Deck as an example:
We need more Linux devices like this to gain market share.
You got it. The moment you surface the idea that there are multiple distros or DEs you've missed the goal the thread is suggesting. Presintalled, customized software built for the hardware is the way to ease people in with zero tweaking, which is crucial for newcomers.
I think this was Steve Jobs' primary skill. He could see a clear vision of the product people didn't know they wanted. Bottom to top, from the hardware to run on, to the typeface their apps used; he knew that the best user experiences happened when every level of the stack harmonized to create a very finely tuned user experience.
Unfortunately, the people who are that good usually don't work for free. We're very fortunate that Valve is choosing to open source their work and keep their SteamDeck platform an open one.
He shipped enough clunkers (and terrible design decisions) that I never bought the mythification of Jobs.
In any case, the Deck is a different beast. For one, it's the second attempt. Remember Steam Machines? But also, it's very much an iteration on pre-existing products where its biggest asset is pushing having an endless budget and first party control of the platform to use scale for a pricing advantage.
It does prove that the system itself is not the problem, in case we hadn't picked up on that with Android and ChromeOS. The issue is having a do-everything free system where some of the do-everything requires you to intervene. That's not how most people use Windows (or Android, or ChromeOS), and it's definitely not how you use any part of SteamOS unless you want to tinker past the official support, either. That's the big lesson, I think. Valve isn't even trying to push Linux, beyond their Microsoft blood feud. As with Google, it's just a convenient stepping stone in their product design.
What the mainline Linux developer community can learn from it, IMO, is that for onboarding coupling the software and hardware very closely is important and Linux should find a way to do that on more product categories, even if it is by partnering with manufacturers that won't do it themselves.
System76 is doing that these days. They put extra hardware support for their Linux distro TuxedoOS and I've heard good things.
Edit: System76 make PopOS and Tuxedo computers make TuxedoOS
I think you meant Pop!_OS (is developed by System76). TuxedoOS is developed by Tuxedo Computers, which is a European Linux focused hardware company.
That said, the point stands... there are hardware companies making Linux supported devices.
Underlying kernel aside, I think that the Steamdeck's SteamOS is an excellent example of how "easy to use" != "smaller feature-set". I've heard countless times from apple dudes that the reason that their stuff allegedly "just works" is because of the lack of some functionally that if present would overwhelm the user. You know, as if ios and android don't share fundamentally the same user interface principles. But they do have a point, a green user can be overwhelmed when presented with a huge feature set all at once. Yet, despite SteamOS literally having a full-blown desktop environment, the UI frankly is way less confusing than my Xbox. It just goes to show that it's not about the number of features, it's about how they're presented. Power users don't mind digging into a (well designed) settings menu to enable some advanced functionality, and keeping those advanced features and settings (with reasonable defaults) hidden around the corner behind an unlocked door helps the newbie get started with confidence.
To make Linux more appealing to the average person, you'd have to be able to buy a Linux PC at your local computer store. Most people can't be bothered to install a new OS.
I’m a very casual Linux user and in my experience, I’ve NEVER had a problem with a documented solution that didn’t require going down a rabbit hole of other references.
Something like this: “To get the trackpad to work with Ubuntu, make sure you’ve installed the hergelbergelXX package.” (No link, find it on your own!)
Visit the HergelBergelXX page. To install Hergelbergel on Ubuntu, you must install the framisPortistan Package Manager. (No link!)
On the FramisPortistan GitHub readme, we discover it requires the JUJU3 database system to be installed. “JUJU3 may cause conflicts with installed USB devices under Ubuntu” JUJU2, which shipped with Ubuntu, is no longer supported. Also we recommend Archie&Jughead Linux over other distributions.
And this essentially never stops.
All of this is comparatively a happy result—I actually DID post a question on linuxnoobs about getting my trackpad to work with Ubuntu… and have not had a single reply. I have no idea how to find out how to make it work.
I had similar stories getting Wireless Networking to work on some devices before. Good thing is, there are drivers for most, if not all, default hardware interfaces directly in the kernel nowadays and if a device has any sort of popularity it will be supported before long if it isn't out of the box.
Hat a problem with WLAN on a laptop when I tried to install fedora. The solution was to install Linux mint with LAN\internet and let the driver manager figure it all out.
Maybe that helps.
Simple, start teaching it in elementary school all the way up through high school. Apple did it long ago and got apple users out of those kids. Microsoft does it now, and now you have Windows users. Just need the computer education to be Linux centric from the start. It's not that it's different, it's that it's not what they grew up with and were taught.
Atomic OSes should be evangelized more aggressively to laypersons. IMO, they’re great for 3 specific use cases:
Need hardware with it pre installed with a reason to buy other than because it has Linux
Maybe use the lack of a requirement for a Windows license to bring the price to performance ratio down
If they're really performant machines also helps break the idea Linux is only for old and slow machines, I only ever used to put it on laptops as they were reaching the end of their usefulness, the moment I put it on my pc and a new laptop it changed my perception on it entirely
I also think the majority of technical users still use windows, maybe we should concentrate on getting them first and maybe we'll see more support
On top of being preinstalled, we also need google search-able instructions that avoid the terminal altogether. People are afraid of the terminal, it doesn't matter why, it just is.
Currently, most solutions to linux problems come in the form of terminal commands. We would have to start creating a whole new troubleshooting forum where instructions avoid the terminal and are just lists of buttons to press in a GUI. Probably helpful screenshots too.
Of course I have no idea if some things even have GUIs at all, like configuring user groups and permissions or firewall settings, someone would need to make them. Not to mention every DE or program would need a different set of instructions, GNOME or KDE, firewalld or iptables. It'll be a lot of work.
I’m tech literate and use the command line daily. I enjoy how powerful it is but I also enjoy the ease of point and click on windows.
After a hard day coding at work I much prefer poking around windows than using a command line on Linux.
you can't because it's explicitly against the whole point of having endless choices. when everyone works on something different, the quality spreads out to where it's mostly just mediocre stuff across the board.
hardware compatibility is also a huge problem. for everyone that says "it works fine for me" there are a thousand others for whom it does not.
I feel like there's also the point that on Mac OS a lot of stuff "just works" because everything else just doesn't work at all. I have a number of things that just aren't going to work at all on Mac. Linux is obviously much more permissive, which leads to a lot more kinda working stuff that just wouldn't work at all on Mac.
Needs to be pre installed, most people don't know how to reset their PC, let alone install a new OS.
It's current year, I should never have to touch the terminal for anything. I don't care that it's powerful, my brain is already full of windows knowledge and I don't want to have to google what command I need to perform basic functions. Everything needs guis. If there's a gui, I can figure it out and also discover tools I didn't know about along the way, which allows me to solve future problems without going insane.
That's popular sentiment though, so how about one that I don't see often: Add options to allow windows like behavior. For example, middle click paste is the bane of my existence. I should be able to change it to middle click scroll os wide, not just in firefox. I know that there's a hacky workaround to kinda make it work, but it sucks.
More GUI front ends for stuff. This takes away the need to understand command line tools and syntax, and makes the out-of-the-box experience feel more like it just works.
Most people have had great answers coming from the company side of things. I'll take it from the standpoint of individuals like us helping someone linux curious see the light, while still having the "just works" experience.
Do not give them any choices. None. Put them on your stable distro of choice for a new user, call whatever that is "Linux", and be on your way.
But why? Isn't that antithetical to everything we value? Yes and no. We value choice almost above anything else, but that doesn't "just work" for most people. Which of those do you value more?
Thank you thank you thank you for posing this question.
This is the biggest issue by far with open source stuff in general, and as a non-programmer who wants to use more and more of it, user unfriendliness hamstrings so much.
I don't know the answers but I can tell you for a fact that if open source in general is serious about broader adoption, this needs to be occupying 50% of everybody's open source discussion time, at least.
What I know is the standard "fuck you read my 19 pages of 1s and 0s" is the wrong answer.
Maybe good design is just really hard. I don't know, I've never tried to do it. Seems like the sort of thing that might take three thousands iterations.
Whether any OS could ever just work isn’t even going to solve the issue.
Getting OEMs to sell laptops and desktops in Best Buy (or the like) that have Linux installed and is properly supported — that is what will help solve the issue.
Software, 1,000%. I love linux and daily drive it. But when I have videos to edit, photos to rework, or collateral to design I have a windows laptop with professional grade tools to do the job.
I'm sorry, gimp is hot garbage. There isn't a pro-grade, open source video editing tool or anything close. Inkscape is useable in a pinch. Scribus is useless.
Not everyone is a multimedia creative professional, but most software on linux never quite have the features you need, are no longer maintained, or will be useful in ten years.
That said, I'd still rather break out the laptop when doing client work than daily drive MacOS or Windows 11. Either way the barrier for most users is that linux almost works.
The way to get Linux more appealing is to get proprietary software makers, like Adobe, Microsoft (Office), you know the actual things people need to do their job, to make software for Linux. Steam Deck is a good example of this, it works because Steam ported the games to Linux...
I think it should be: "Software that is yours"
Overall, I think more focus should be put on consolidating similar projects.
Do we really need 6 different window managers that follow the same design logic?
Do we really need each major distro to have its own package manager?
How many image and PDF viewers do we need? How many music players?
Can we convince Ubuntu that no one wants snaps and they are wasting developer resources.
The freed up capacity should be focused on better windows app compatibility. Something akin to Valve's push in gaming.
While I understand the sentiment, we have to understand that Open Source developers work on projects that motivate them.
So, we can have a single example of each of these but they do necessarily get any more devs. In fact, if you take economic theory ( competition for example ), it is likely they attract less attention individually than they do competing as part of an ecosystem.
It would certainly help on the user acceptance and commercial software side where choice is an impediment. But, if we are just talking resources, limiting the number of projects only works if you pay people to work on them.
Why was each of these projects started ( eg. window managers )? The answer is simple. It is because the founding developer did not like any of the existing options.
As others have said, macOS does not “just work” anymore.
I am primary tech support for a few “normy” users including my mother and wife. My wife, the more technical and capable of the two, uses macOS. My mother uses Windows. My wife requires substantially more tech support. Worse, the issues are often complete mysteries to me like “why is everything so slow” and it turning out that some OS level process is consuming huge amounts of memory and / or CPU. Web searches reveal lots of people with similar issues but no real insight into what to do about it or why it is happening. I have moved OS versions just to solve this kind of crap on Mac. Another problem is software not working on older versions of the OS.
I am no Windows lover but, once I show my mother how to do something, I never hear from her. Every once in a while I stop by to marvel at how many updates need to be applied but that is about it. She is in the Windows 10 that I installed for her many years ago now. It just works.
From a non techy perspective, having what is used and installed being secure is a big one. I am new to Linux Mint. Mostly user friendly until something gets corrupted or suddenly can not be verified.
Looking for why is not always simple, and there are some explanations/instructions easier to understand than others.
I preface most of my searches with Linux mint (whatever I am searching for) for dummies. This helps some.
I have been forced to use mac now for like a year, and I don't get the whole "just works" opinion of it. Like I have had so many issues with just basic stuff. Turning off mouse acceleration and the mouse still feels all slimy. Highest mouse speed is so slow and setting it higher requires some crazy tricks, which also does not work consistently through boots. It can't wake up a lot of monitors, I have to turn them off and on manually. If it cannot connect to a monitor properly but tries, it like disables your keyboard for a few seconds while trying. Some items in the settings menu take a long time to load, as in if I reboot, log in, open settings, there is no mouse settings.
Software-wise, it seems that the relatively fast adoption of flatpaks and other containerized formats somewhat solves the typical dependency hell that was so common in Linux just a few years back (and to some extent still is an issue today depending on your distro and use case). The hardware support side is a little harder. That's going to be up to vendors to play nice with the Kernel team and/or introduce reasonable userland software that doesn't break the golden rule. Until Linux gets more market share the latter isn't likely to happen. A nice side benefit of the emergence of immutable and/or atomic distros is that users can play around and try things with much lower risk of bricking their systems, so I'd also consider that a step closer in the "it just works" department.
Honestly Linux does work pretty much just as well as MacOS if you run it on hardware thats super well supported and that tons of linux users use. MacOS has integration with its hardware because its all made by the same company. They only have to support a few models of computer.
If you installed Linux Mint today on a Thinkpad t480, and on some obscure weird laptop with rarely used hardware your gonna get 1 install that just works out of the box and your gonna get 1 that you have to hunt for drivers, and do tons of work on. Its just the nature of being able to use any hardware. Some will work better than others.
If you want an example of how to increase adoption you pick a line of computers thats of high quality and have them be supported by the community a ton. Then you convince the company that makes these computers to ship a version of them with linux pre-installed, and potentially help atleast with funding the development of whatever distro they use.
If your average user bought a laptop, opened it, turned it on, and it had linux on it and worked relatively well, they are never going to change it. Its not a normal thing to just change your OS most people don't even know that you can do that. I gave my grandmother a linux mint laptop and she thinks its windows.
The issue starts at the fact that it's difficult to find a computer sold by a common major distributor with Linux already installed, nor does Linux have any marketing aside from word of mouth to compete with the aggressive Microsoft/Apple duopoly.
The threshold to entry begins at simply having the technical prowess to install an alternative operating system on one's computer, which I don't believe a good majority of people are even capable of. Before that, people also need an incentive to transition in the first place. They've probably been using their current OS for a good portion of their life and are more than comfortable with it without putting themselves through another learning curve.
The average person isn't considering an alternative to what they're already using, and if they are, it usually isn't Linux. The biggest problem isn't appeal or ease of use; it's exposure and immediate accessibility.
That said, performance and simplicity would be an excellent selling point for Linux. It would be absolutely worth banking on the open-source nature of it to appeal to a growing demographic of people interested in privacy-oriented tech as well.
Kinda don't think you can its one of the beauties of Linux, there's so many different flavors of it. Best thing that would've helped me as a beginner would've been like a collection of all the wiki's and basic knowledge in a single space instead of searching through different sites for a problem or terminal commands, which I bet exists but I just never looked too hard. Also documentation of common problems would've been big for me (especially for older devices) like drivers no longer being supported by kernels and solutions like using the open source version instead.
The problem is, that no operating system "just works". It also highly depend on what the person wants to achieve, and if there are any pre experience with computers or even relying on existing software or specific hardware. My recommendation is not to tell people the illusion of "just works" and be honest upfront. People should learn how it works, what to expect and if tradeoffs, time and resources are worth it.
Same is true for the other way too. Does Windows "just works"? Especially if someone switches from Linux to Windows.
Rather, we should teach the reasons to switch and encourage that decision. In example why it matters to have control over your system, rather than the company has control over it (MacOS and Windows) or why spying on you is bad (Windows). And encourage giving up something you are used to (and maybe paid). Sometimes its okay to use a program that is not as good as Photoshop. Sometimes its okay to give up playing a videogame you like (and maybe associated with friends playing that game with you). But most people are not ready to do it, because that is associated with lowering quality of life.
I switched in 2008 from Windows XP to Ubuntu. I know these struggles. And they are not over yet. This is an ongoing task between my brother and me too, and he was using the Steam Deck, but decided to go with Windows 11 with the recent build. It was almost there, but there is always a butt. I say, don't tell people that "Linux just works". No operating system "just works".
Linux is a tool that big corporate entities have profited greatly from for many years, and will continue to. Same with BSD, Apache, Docker, MySQL, Postgres, SSH...
Valve, Sys76, Framework, etc. Are proving that using Linux to serve an end user market is also profitable, and are capable of supporting enterprise use-cases.
I understand that there may be specific problems to solve wrt improving adoptability, usability, compatibility, etc., but Linux is doing more than ok within the context of the FOSS ecosystem (and increasingly without).
Your thinking is slightly skewed, IMHO. Linux doesn't have an inherent incentive to compete with MacOS or MS, and if it did, it would be subject to the same pressures that encourage bad behavior like spying on users, creating walled gardens, and so forth.
MacOS being a bad example here since Apple only needs to make its OS work on a very small set of hardware that they control wheras Linux (and Windows, yes) need to work on probably hundreds of thousand if not millions of devices (including Macs 👌) with at least the same amount of peripherals combined in almost any imaginable way. That's a completely different task.
From Wikipedia, the free encyclopedia
Linux is a family of open source Unix-like operating systems based on the Linux kernel, an operating system kernel first released on September 17, 1991 by Linus Torvalds. Linux is typically packaged in a Linux distribution (or distro for short).
Distributions include the Linux kernel and supporting system software and libraries, many of which are provided by the GNU Project. Many Linux distributions use the word "Linux" in their name, but the Free Software Foundation uses the name GNU/Linux to emphasize the importance of GNU software, causing some controversy.
Community icon by Alpár-Etele Méder, licensed under CC BY 3.0