Ask HN: Main reason that year of the Linux desktop hasn't happened (yet)? | Hacker News Reader

Ask HN: Main reason that year of the Linux desktop hasn't happened (yet)?

10 pointsURfejk17 comments

I installed Debian (supposedly a distribution that is stability-oriented) just today. After initial restart, it greeted me with... a blinking cursor (not even a prompt) in the upper-right corner of the screen. Knowing Linux, I knew that I need to Alt+F1 into a terminal, log in and then troubleshoot what went wrong. Turns out you need to manually download and install some ATI firmware or the X won't work. I wasn't informed of that anywhere in the installation process. As long as shit like that is commonplace, Linux adoption will be limited to fans and IT specialists.

BTW once I fixed that, it turns out that Xorg, even with the recommended drivers, is just too crappy to maintain Vsync and everything (even just moving of a window on the desktop) is slightly, but noticeably tearing. As a remedy, I switched to Wayland, which is designed to not have this issue, and everything is butter-smooth. Why is Debian shipping with an option to install the crappy, tearing windowing system in 2020? Many users will, not knowing about the Wayland fix, want to go back to Windows after just couple hours of use.

Debian is stability oriented not user oriented. Ubuntu and Fedora have better defaults and work well.

This question presumes that "the linux desktop hasn't happened". Something that is very much a definition issue.

* What is "happened"? Millions of people use Linux on their desktop. Probably more people (absolute numbers) use Linux today, than windows3.11 back then. Does that mean windows3.11 never "happened"?

* What is "Linux desktop"? Is chromebook a Linux desktop? A raspberry-pie? A laptop? The terminal on the counter of some large bank or PoS in a large shop?

What also strikes me in this, is that it used to be "the year of linux", until Linux emerged in every niche as clear "winner" (firmware, mobile, terminals, servers, supercomputers), at which moment the meme pivoted from "year of Linux" into "year of the Linux Desktop".

I'm not saying OP is doing this, but I have a strong feeling the phrase "year of Linux on the desktop" is mostly used by people who want to show that Linux is failing.

Edit: which it clearly is not: Linux is Open Source. Open Source has "won" the very moment one person is using it successfully. The fact that I'm typing this on my Linux machine means that Linux "won"; don't forget that Open Source software has no- or at least a very different business model than most closed-source software: success is not (nessecarily) defined by "market share".

There are too many Linux distributions and too many desktop environments. There is no coherent message as to what software to use. It is difficult to develop a critical mass of users that uses the Linux desktop.

Lenovo announced recently that they will support Ubuntu in their range of P-series laptops because they are interested in the near 3% total Linux global desktop users.

Linux desktop is simply unreliable. I know this might not be what the community would like to see in the comments, but in my experience, the linux desktop has a long way to go in terms of reliability.

For example, an ubuntu update, literally crashed my system and i managed to recover my machine from collapse. I am a techie, so i managed to do this, how many normal people can do this ? Would they know where to look for the issue ?

Also, the desktop freezing, and constant popups asking to submit reports for some internal issues is quite an annoyance.

Having said these, using only cli linux is the best and no OS comes closer to the versatility of linux. But as a desktop, i feel its has not made it. Reliability is a core feature.

It's not good enough for most users.

This has been my (unfortunate) experience trying to help some friends and family over to various distros over the years. Linux is in the odd cycle of assuming users are comfortable using the command line to do things which inevitably leads to users who aren't comfortable not using Linux, forming a self-re-enforcing cycle. The entire Unix design philosophy of small tools that are chained together that (most) know and love only really makes sense under a cli first approach because we don't really have a semantic to pipe and redirect output in a graphical environment.

Main reason is the lack of a holistic approach. Every Linux desktop is a loose collection of separate programs that don’t really play together in harmony. There are no Product Managers behind the distros, it would seem, or if there are, they haven’t been very effective. KDE, GNOME, Cinnamon, all of them scream, “we haven’t done user studies.”

I suspect this is a function of the distributed ownership. No one is able to manage the full stack because the complexity of doing that would be a full time job and take a team working for months to get it ready.

Another way to look at it: what makes Apple successful is at odds with how the components of the Linux desktop come together/are built.

In order to provide for a consistent, interoperable experience across the stack, you need control of its components. Apple's vertical integration (now making chips!) gives it the ability to make changes in order to provide the 'best' experience (best as in, profit; taking into account consumers express their opinion with their $).

The Linux desktop optimizes for freedom: you can choose what parts you want. That's not conducive to a consistent experience.

One could argue that's not the point. If you want 'the desktop' just buy a Mac. If you want 'the Linux desktop', it's been there for a while now, just go ahead and glue together the parts - that's the point!

The freedom of choice for individual components is certainly a benefit of the platform, but I would argue that it also doesn't preclude having a default preset of components that form a cohesive vertical experience. In other words, the defaults should work. That most likely means forking select components, which is what Mint did with Cinnamon, to create a micro-ecosystem that is internally consistent, but they just haven't gone far enough yet.

You can't design a desktop GUI by committee. But when it comes to Linux everyone wants a say. The model for open source code contributions just doesn't work for visual and interaction design. (Or if it does, I've yet to see a successful example that wasn't an exception rather than the rule).

When there are too many participants in the visual and interaction design of a program or OS, you end up with a project pulled in every direction and pleasing to no-one. But if you go the opposite route and limit design decisions to a dedicated UX team (like Canonical did for Ubuntu), you end up generating resentment from contributors or users who feel their input is being ignored.

One of the best Linux desktop GUIs was Corel Linux which was released in 1999(!). Corel understood that a consistent look-and-feel and ease-of-use were essential to attract new users. But even back then they faced criticism from Linux developers over various design decisions. Rinse and repeat this scenario for every attempt to create a consistent, cohesive Linux desktop over the years.

Perhaps too much choice. The lack of a single dominant leader in the desktop Linux distribution space makes standardization a challenge. Thus every Linux desktop is a little different.

Personally I find PopOS to be the perfect blend of what I want in a desktop Linux system, but because desktop Linux isn't popular enough it still lacks support from major software producers such as Adobe.

IMHO, it is because desktop development is:

1. Fragmented. There are probably an adequate number of developers working on the Linux desktop, but they're working on different desktop environments.

This isn't necessarily a bad thing, because nearly everyone can find something that suits their tastes.

2. Not funded. There is not enough financial incentive to get something that can truly, truly compete with the likes of macOS and Windows. The closest, I guess, would be ChromeOS, but AFAIK the needed desktop components aren't open sourced in a usable manner.

Note: I primarily run either Ubuntu GNOME, XFCE, or Manjaro as the primary OS on my home PC -- depending on my mood of the week.

Macs come preinstalled with MacOS, anything else comes preinstalled with Windows. Very few people will even consider their OS as something they can choose, many have no idea what Linux is and the percentage of people who will take the time and effort to switch to Linux is even smaller. So the Linux desktop will remain a thing mostly for people who work in software.

It already happened for me, I've been using it with all my computers since 2000.

I'm curious what would have to happen for you to consider "the year of the linux desktop" to have happened? I'd say it arguably already has happened.

I'm curious why say that it has already happened? Linux is generally reported to only have 1-2% market share for personal computers.

Reply on news.ycombinator.com