Picture taken from wikipedia.
There are issues when it comes to Linux being more broadly adopted on the desktop. What are the main issues you think and how can they be addressed?
I think that the big issue with Linux is fragmentation. This has been especially an issue when it comes to developers publishing games on Linux.
The solution to this issue are cross-distribution installers like Snap, AppImage and Flatpak. Every application has its own set of libraries, meaning they do not depend on distribution-specific packages.
Another issue I believe is momentum. There's some outdated opinion on how well Linux works on the desktop.
Defeating the momentum will require more effort on the PR front. I don't see this being something that will just fix itself with time.
Thoughts?
Oldest comments (102)
I’d guess that the people most involved in and attracted to Linux originally didn’t have the sort of design taste, non-technical user empathy, and marketing savvy to make Linux desktop the standard.
Linux grew in a way that reflected its originators and eventually lost too much momentum on the desktop side of things. With tons of success in extra areas, Linux folks started leaning more in on strengths and desktop became more of a niche.
With open source becoming more mainstream and desktop computing stagnating of late, I think there’s room for the next great OS to be open source, and it would be great if it were more grass roots and not a Google project.
Do you think a grassroots project could compete with OS X and Windows?
I'd love to see it done, but I think it would need strong sponsorship from at least one major player in the software industry.
Could, yes. It's not the most outrageous idea.
Tough barriers. Maybe not totally grassroots, but a startup. What if it were a startup that built open source software and served a global community of developers, a DEV Community you might say. 😅
I'm just kidding—sort of. It's a project I'd love to be a part of if the stars lined up in any way.
There are some technically cool new OS projects like Redox. Oddly I don't have a lot of expectations that it will succeed greatly, but I'd also get pretty excited about the prospect of personally being involved in making that happen.
How is that different from Canonical? Ubuntu was grass roots, and they invested a lot into the desktop experience. Also, I think modern distributions don't reflect that era of engineer-minded design anymore (unless that's their explicit goal).
I think you’re right. But I think the history still matters, just like how Apple’s founding history and Steve Jobs still help sell iPhones.
It shouldn’t matter as much as it does, but the world sometimes contorts itself to make it matter.
I think the biggest issue is UX for non-technical users. apt-get or dnf might seem totally fine to us, but if a non-technical user has to open a text interface to complete a mundane task you can forget about it.
Some projects are working on it, but without some big changes and a lot of marketing Linux will probably always be a niche tool. There is nothing wrong with that if it works great for its niche.
So far I haven't had to use apt-get to install GUI software (only need it for some development stuff). Everything can be installed from Ubuntu Software (it integrates snaps, gnome plugins, and apt packages). I'd like to hear what software can't be installed through a GUI.
I personally want to have Linux on the desktop become more mainstream. Right now at work I'm forced to use OSX. I want to use Linux as my only operating system, but due to it not being mainstream enough it doesn't look like its going to happen any time soon.
Where I work people choose their preferred OS and we haven't ever had problems, not sure why that's not more common practice.
I've worked in enterprise before and I can see the struggle with that in those environments, but I feel like devs should be encouraged to work in their favorite OS... 🤷♂️
Can't speak to "enterprise" per se but our office is split 50/50 macOS and Linux and we haven't had problems getting work done.
I was sat in a workshop at an enterprise that has a fifth of a million employees and one of them had a mac book pro. I stared at it a lot as the company has a few hundred thousand IBM thinkpads with the majority running windows7. On the coffee break I ask him if it was indeed a company laptop on the global corporate network. He said it was. The developers at this enterprise are given Thinkpads. So I asked him what he did for a job. He is a data scientist. I guessed that the head of data scientist had argued that they needs macs for some made up reason and the enterprise had gone along with it as a special case as it's very hard to hire data scientist and impossible to retain them if you give them a winows7 thinkpad on their first day 😂
Not only apt-get. I'm not a geek but I used computers in the 90s. I still cannot use midi on Linux and I've tried for hours and read much info. Some things are difficult to set for someone who just uses the computer as a tool for other things.
Sure apt-get is just an example. Many things aren't done with non-technical or casual users in mind in the Linux ecosystem.
This, in a nutshell. I am a technical user. I've used Unix and an various CLIs, but every time I see a set of instructions that tell me to use apt-get my heart sinks. Give me Windows every time. Download, double click to run - OK - OK - OK - done. And yes I installed a Linux distro recently, and had to run apt-get multiple times to get the software I needed.
I ask myself this as well. I had used Ubuntu for quite some time and loved the fluidity and customization - not to mention it being open-source. I'm convinced more people would convert (especially from Windows) if they gave it a shot.
The biggest reason IMO are apps not having a Linux version. Users are hard-pressed to switch to Linux if they can't run their Windows app.
What the world needs is some sort of container for desktop applications. Developers should be able to code in whatever language they want and ship it as cross-platform by default.
The OS vendors should team up and make a standard API to enable containerized desktop apps. This way the desktop world can stay afloat in the cloud-native era, where webapps are becoming more capable than before.
So, Electron?
Hope not. Imagine having 10-15 apps all self contained Chrome browsers :D
Maybe if they shared the same runtime... like an operating system ;-)
Correct me if I'm wrong but AFAIK Electron is JS/HTML/CSS only. Meaning you can't do apps that require special runtimes/engines/dependencies e.g., Photoshop / Video editors / Advanced 3D games or anything that requires a native desktop?
What I think is missing is a "Docker" for these kinds of apps.
Millions of people still want to use "next, next" buttons. Let's assume 3 billion people using computers. Maybe all developers' count is 15 Mio.
The other people might don't want to understand technical things. Because of this, most people using an easily understandable operating system.
Although I don't like Windows, the Windows offers an operating system which even most stupid people can easily understand.
I'm using Linux since 2007. I was starting with Pardus. It was very complex for the newbies. I had a hard time installing Pardus but I could install Windows myself.
And comments are saying, the design isn't good. There are different window managers. Nothing has changed during all these years. This shouldn't be called stability.
unexpected imo here:
I guess it is because lower computer literacy in modern society created by corporations and education system. While apt-get/dnf thought by @jacobherrington is true, it is just result of this. Anybody remember DOS/Home Computers era and how much time MS and others spent to push GUI as desktop OS? Main problem is people don't want to learn how to use their products anymore and IT helps them with it.
I think that might be selection bias, 20-30 years ago everyone wasn't using computers. And the average person was using computers for very different tasks than today.
Sure, not everyone, but nobody thought about computer like about "don't need to read instructions" thing. Level of illiteracy on mass market is like "oh my god! there is no button labeled Start! only mystical Begin! What will I do!". People use computers and devices mindlessly and nearly panicking if got some text message or instructions. While it is good for "sellers" it is disaster for society and communities.
I'd disagree. I think we just need to know who we are writing software for, which is why I'm okay with Linux serving a niche.
I don't like Windows, but it has enabled millions to have access to information like humanity has never seen before.
Empathy is a big part of this discussion.
Sorry, but IMO, not Windows provided access to information, it was lowering of hardware prices. There are already was fair amount of different OSes before and in same time with Windows. In many countries in schools and even some colleges/universities/etc "computer science" was (and still) synonymous to Windows + Office button remembering instead of teaching concepts of computing.
Fair enough. It's hard to point at one cause of the explosion in computing, and you're definitely right about hardware prices.
Linus answered this question in 2011
2012
To answer the question, first look at why Windows and OS X are as popular as they are currently.
Windows comes with most desktops or laptops you might buy today. To get hardware without Windows, you have to go out of your way. Dell makes machines they'll put Ubuntu on, but it's not easy to find those, and you really have to be looking for it -- which is not what most customers are going to do.
OS X comes on Macs. Most of those buyers want a Mac for whatever reason, and they get OS X with it. I'm an outlier here because one of the main things I like about Macs is they run OS X.
Most consumers aren't out there going "ok, so I need a machine that runs OS X on it, or even I need a machine I can install OS X on it." They get whatever OS comes with the hardware they bought.
There are companies selling Linux machines called "Chromebooks" :) . Again, they're not looking for a Linux machine -- or even a machine that runs ChromeOS necessarily. They want a cheap machine that's secure and can run stuff in a browser.
I think Chromebooks and even Android answer the question here. The only way for "the year of Linux on the desktop" to arrive is for some company to make a device people want to buy that just so happens to run Linux. Most buyers out there don't care what the OS is as long as it does what they need it to do, like run MS Office, run the games they play, etc.
I think macOS is at this time, the best for an all purpose machine. You have access to proprietary softwares (Adobe suite, MS Office) and FOSS sotfware (homebrew, docker). macOS is great and somewhat intuitive operating system.
I post that, if Adobe can run on linux, i migrate immediatlly to Linux here.
Macs are expensive machines.
More and more expensive and less and less customizable, unfortunatly.
I think Louis Rossmann would have a differing opinion.
He doesn't fix operating systems either.
There are actually other companies selling Linux machines such as Dell, System76, and Purism. I do get your point though, not much that's geared towards general users (most of these options target developers in their marketing).
Here in India personal computers became popular during the late 90s, and sellers used to install pirated copies of Windows 95/98 when they deliver computers. During the 2000s the personal computer market grew exponentially, and most of them came with Windows XP --- either OEM installed or pirated.
During that time, Microsoft was raiding offices and institutions searching for pirated Windows, but personal computers were never checked, thus making Windows the number one choice among people. Many applications were built for Windows.
People got used to the Windows UX very much. The UX remained pretty much same from Windows 95, 98, XP, ME, Vista, 7, to 10.
They know that "My Computer" shows what's in their HDD. The letters "C:", "D:" are their HDD partitions. Inserting a CD/DVD and playing music out of it was very easy. They know how to download something and install, how to change the wallpaper, how to change the default music player. Adding a new printer or setting up a network connection was so easy.
Linux was fragmented too much at this time. Also, their UX was not standard across distributions.
If a normal user wanted to try a Linux distro, they started facing many issues.
First of all, installing Linux on a desktop was not easy for a normal user. Then, people couldn't understand the file system of Linux. "/home" was a mystery for people that came from "C:", "D:". People had no idea where to look for HDD partitions. They didn't know how to install an application. Downloading source code and compiling it was not fun for a normal user. Doing something in the command line was cumbersome. Adding a printer/scanner was not easy. Audio may fail because of a driver error. Ooops!
Within days, they would roll back to Windows. I mean, their pirated copy of Windows.
Today, there are better Linux distros with much improved UX. Now there are good office packages, music/video players, and Chrome/Mozilla --- which are the primary apps used by a normal user. Many computers comes with Ubuntu OEM installed. People are fine with that as long as their Chrome/Mozilla works and music/video plays well. GUI based package managers are there, so people are fine with that too. Reading a CD/DVD or adding a printer are easier now. Anyway confusion on the "/home" filesystem still exists among them.
Today a good percentage of laptops come with Windows 10 OEM installed and a few models with Ubuntu (or some other distro) installed.
The popularity of Linux (primarily Ubuntu) is increasing slowly, but it can't dethrone Windows in a near future for sure.
You nailed it.
The reason is very easy: Microsoft Exchange and the related ecosystem. It makes hard as hell to introduce linux desktops inside companies. So people at work will always install Windows, and maybe Linux in some virtual machine.
I would say the lack of PR, paired with Microsoft's historic smear campaign (which they appear to have quit), is half to blame for the lack of Linux popularity on the desktop. It certainly has nothing to do with ease-of-use!
I've introduced more average computer users, especially those considered "computer illiterate," to Ubuntu, and most of them have raved about it to all their friends, with them detailing how much superior it is to Windows in their estimation. It takes me 15 minutes to get an average user comfortable in Ubuntu, and they very rarely need to contact me for support. (Once in a blue moon, I get a phone call.)
I'd say the other half of the problem is the big software companies refusing to offer support for Linux. Adobe, Autodesk, Avid, Ableton, Intuit, and (decreasingly) Microsoft have waged a war against the platform for years. Thus, the main obstacle to adoption for many users is "my software doesn't work there." There are two dimensions to this:
Protools (Avid) and Ableton, as well as some Adobe and Autodesk products, sport workflows that their users are well-accustomed to. Typically, however, this can be overcome once they find open source alternatives that have the same features.
The larger complaint I hear is "I can't open/save the file formats I need." Intuit, Adobe, and Autodesk are especially infamous for this, so much so that I wrote an article calling out their products as ransomware.
Of course, the only reason those products are so ubiquitous is because users are told they're ubiquitous. For example, the majority of schools and universities push Adobe and Autodesk products, even though there are many studios and companies that refuse to use those products, many preferring in-house or open-source tools. And ultimately, the only reason the schools push the products is because they're told those are the standards...and the only reason those are allegedly the standards are because the companies in question claim to be so.
I could go on, of course, about how any decent educational program will teach workflow-agnostic skills, so the student becomes proficient with ANY related software...but I digress.
In the end, then, Linux and open source aren't more ubiquitous primarily because the competition saturates all marketing channels with claims of being "the standard. The only way to overcome this is to do what the open source world has been doing for decades: educating one person at a time.
So true and let's not forget about different "near bribe" projects and sponsorship of schools from big companies that helps them to choose "industry standard"
Yes , that's true.
How much can you save in a company only buying one soft like, for example, Adobe Master Collection? you won't need a linux license per pc.
Us developers weighing in on why Linux isn't mainstream on the desktop is one the reasons why Linux is not mainstream on the desktop :D
Others are: historically ugly and inconsistent UX, the fact there isn't a single Linux but many distributions, Windows being pre-installed on millions of machines, years of Linux not being really ready with people blindingly say "it's ready" (and they were all devs), the existence of Apple, OEM deals, millions of dollars spent on UX research in the other two major OSes, major apps not being available (Windows Phone is dead because apps weren't available and devs didn't want to code for Windows Phone), compatibility of formats.
I don't think there's a single reason why Linux is not more popular on desktop.
I think the extreme diversity of distros is what fundamentally ruins Linux's experience on the Desktop.
For most people, it's either Windows or a Mac. It's usually that simple, and makes for a relatively easy choice when deciding on an operating system.
Linux however, doesn't fall under this. It's a kernel, which is built upon by others to create operating systems. The problem any new linux user finds when attempting to join is, which to use? People will commonly say things like 'Ubuntu', but to an end-user, that doesn't mean anything. They still don't understand why there are so many choices. It confuses them, leads them into problematic situations of choice, learning, and other horrors.
Think about it. If you have an issue with your OS, usually you'd simply search up 'Issue with Windows' or Mac or such, but when it comes to linux, no such situations are universal. When can
sudo apt-getbe used and when issudo pacman -Sused?The vast choices also make it hard to try it out. You have choices from things like Debian, Arch, Fedora, OpenSUSE, etc etc etc. What does the user choose? What is easiest to work with, starting as a noob.
I think fundamentally, the main problem ends up how diverse distros are. It's so unclear and confusing where a new user should go to learn stuff that they simply stop trying to enter the ecosystem.
Freedom has it's price when it comes to Desktops unlike phones which seem a lot more personal and deserve more "customization" in which stock android already dictates how that experience should be, I don't see any stock Linux in the same sense there's a stock android.
Agreed! there are many choices!
Perhaps you could contrast your question with another question: why is linux on the server so successful? Viewed from that perspective every reason why linux doesn't work on the desktop seems like a shallow excuse.
Take for example fragmentation: Hell we have servers running on Ubuntu, Debian, Redhat, suse etc. seems no problem here.
The point is: Linux filled on the server side the need of having an alternative operating system which was cheap and tuneable.
The same goes for android.
But do we have the same need for the desktop?
No.
And that for two reasons:
1) Your consumer product comes with a desktop operating system pre-installed. You e. g. gain no monetary advantage if you choose Ubuntu over Windows. So it's more of an ideological choice.
2) Desktop - or should I say - stationary computing is on the decline. Aside: I am writing this with my tablet lying on the sofa.
So my expectations are, that there will be no alternative desktop operating system an the near future.
@ben Halpern
As much as I would like to see a more "grass roots" type of OS,I am well aware that the main part of the success story of linux incorporates names like Redhat, IBM etc. (even ironically Microsoft is nowadays a big contributor) so that I see that that is not going to happen any time soon.
All good points
I wonder if Google Fuchsia will amount to anything in the future...
Fuchsia is a whole new start, and not Linux based. If Google wanted to flex their hardware muscle we could see any of Fuchsia's layers on any number of devices from the Home product line, future would be "chrome" books or Android devices.
I like the layers idea a lot, it means you can pick as much of the core OS as you want and then build on top of it, hopefully, circumventing the traditional fragmentation issues associated with custom Android builds.
EDIT:
I just noticed the Wikipedia article doesn't mention the layers here's 9to5google's take on it 9to5google.com/2018/03/16/fuchsia-...
I am not seeing how Fuchsia is an improvement in terms of the shattered ecosystem Google produces. Reading sentences like
make me sick: If phone makers are able to mix and match everything up to their pleasure, they are just going to do that. Not only comes your next phone with preinstalled crap, it is higly incompatible the rest of the "Fuchsia"-ecosystem.
Coming up next:
App developer's hell
»Your App doesn't work on the recent Fuchsia 1.32.3-HUAWEI-t-mobileEDITION«.
From a consumer's perspective are they doing unwanted marketing for iPhones. Until now, I was not willing to buy wholly into Apple's ecosystem; but dropping Android and coming up with that mixed bag will help.
Think of Fuchsia as a set of building blocks.
They have to always be in that order. If you want to replace the
app-layeryou'd also have to replace theUI-layer.Companies like Samsung want to look distinct but don't want to put in all the effort of making an
app-layer. In Android they replace parts of the code with their new UI meaning at update time we have a stream.But with this new way of doing things we have 2 unrelated update streams.
It might not end up being that smooth, but I'm an optimist 😉
You are missing my point:
Samsung ist going to pee as early in the stream as possible - the same goes for your phone providers in case you are buying a branded phone. You as the customer get only dirt all the way down.
Some comments may only be visible to logged-in visitors. Sign in to view all comments.