Being able to do everything a "normie" would do without the need to use a terminal.

And a way for companies to get flatpak as an alternative, i remember a friend of mine who tried to use 'buntu budgie for a while and he needed a software for cartographic stuff.

We got lucky the company of said software (and yes, it needed to be THAT specific software to avoid compat' issues, so no free alternatives were viable even if they were available) used to provide a .deb package, we got forced to change a lot of sys native binaries to make it work and ended up just breaking a lot of other stuff to do so. Flatpaks (fuck snaps) need to be the default option to be available across systems without caring about distros, so anyone can run it on "mandragora linux" if they want to

I'd agree with the: come preinstalled. Most people buy a device and never change the operating system. So it needs to be the preinstalled operating system on the average computer or laptop, wherever people buy those.

(And mind that Linux completely dominates the market on servers. So technically, a lot of people use Linux in a way... Just not on desktop computers.)

• Needs to come pre-installed on computers.
• Pre-installed distro needs to support one-click installation (like .app or .exe).
• Pre-installed distro needs to have be easily searchable (for problems, and e.g. searching "chrome DISTRO_NAME" needs to pop up with a link to the one-click installer).
• Pre-installed distro needs to run perfectly out-of-the-box, no fiddling with drivers, no needing to issue a random shell command for some random issue.
• UI needs to be intuitive. Probably something like KDE. Could maybe do Elementary or GNOME with dash-to-dock or something.
• Updates should be easy. Ideally apps can self-update or the apps will indicate if they need an update and have a button opening up an updater that can update all your apps/the OS.
• Updates for minor programs need to be hidden/rolled into OS updates. Most people aren't gonna want to see that glibc updated.
• Better management of stuff like VPNs (probably not important for the average user, but e.g. NetworkManager's GUI support is kinda shit).
• If using GNOME, need to have app indicator stuff pre-installed (if I'm being honest, the fact it's not built-in is absurd).
• Needs to come with good basic apps. Some of the default apps included with DEs are kinda shit. There is still no truly good mail client IMO (at least that doesn't look dated AF).

Probably more.

EDIT: Something like Lutris should probably be integrated into the OS. Installing non-Steam games is a minor hassle at the moment IMO.

For non-enterprise users only two things:

1. Zero reasonably priced options for support when things go wrong.
2. Breaking changes caused by updates that make that support necessary.

If my neighbor's Windows or Apple machine breaks they can call Microsoft or Apple, the PC manufacturer or a bunch of different support providers. Microsoft provides free support if one of their updates causes problems.

I can't find any Linux support aimed at home users, only very expensive enterprise support options.

CAD software.

FreecCAD just released it's first full version and it's a pain to use. Back in 2018 somebody said FOSS CAD software was at least ten years behind the big windows commercial software. I think now it's about fifteen behind.

For the vast majority of users Linux is just a worse deal. Only thing that really comes to mind that Linux does that users care about is that it will support that hardware that Windows 11 will leave behind, and even those users will happily just run Windows 10 without updates and if that bites them in the ass then maybe they’ll upgrade or just ask their IT friend to use a bypass to make Windows 11 at least work on their old hardware.

Otherwise, of the things users actually care about, Linux has worse app support to the point that even pro-Linux users would rather dual-boot that lose access to their games and worse hardware support. Linux also has a problem of not being well understood by a lot of tech folk so if you bring somebody onboard you better be ready to be their only point of support.

ChromeOS is probably the best example against this since it is basically just a browser, the laptops it sells on are substantially better value than their budget counterparts and realistically a lot of the people buying them are parents for their kids so the user’s preference is substantially pushed aside in favour of cost. The SteamDeck is another good counter-example since it essentially refuses to compete with the PC gaming market by calling itself a handheld.

Linux is stuck in the crappy position of needing more users to get more software and hardware support but users need better software and hardware support for Linux to make sense compared to Windows. It’s getting better and Valve’s efforts have steadily brought the Linux gaming percentage up but it’s still the enthusiast OS.

By all means encourage it’s usage though. Linux is a far more open and privacy-respecting option and the more tech folk and basic-usage users that adopt it the better!

I think the big thing that everyone is missing here is that schools and workplaces need to push it into people's lives. For that to happen Linux (or at least one of its distros backed by a hardware distributor) needs to develop killer features for those markets and successfully sell to them in large enough numbers that the average computer user - who does not care what their OS is because they only use it for email and work - will make sure that their at-home setup is compatible with their work machine.

That moment is when market forces will take over and drive real growth in desktop Linux, rather than the tiny little bumps we've seen the past few years thanks to the Steam Deck coming out and MS pissing its users off.

This is how Apple built its marketshare against the Microsoft domination of the 90s. For a long time it was the go-to "school computer", and then those kids grew up and now a huge piece of the tech industry and culture is more or less Apple only. It's unclear if this process can be repeated, since Apple's marketshare was carved out during a time of massive growth in the industry that is unlikely to repeat, but I wouldn't say it's impossible if the right conditions reveal themselves.

I will say that it is highly unlikely that the people here would like the change if it happens - imagine Google slinging fully locked down "linux" machines en masse and everybody else needing to download their kernel fork that's loaded with spyware ("for security reasons") in order to connect to Google Teams for work. Maybe I'm being pessimistic but I just don't see mass adoption of a new OS happening without some kind of fuckery like this that renders the version of Linux that gets mass adopted unrecognizable from the version we're all using now.

The other option is state intervention, as with NeoKylin in China, although the Chinese government seems to be limiting themselves to just government computers with that distro.

This question comes up every other week. I reject the premise that "more users" is a commonly held objective.

For most linux / OSS projects the objective is to be the best the project can be. Having an active community is usually part of that but "more users" is a low priority.

This most difficult one is probably the fact that 99% of people do not install their operating system.

The device they purchase needs to have a clean and elegant out of box experience like the Mac. Regular folk who are willing to stray from windows don't consider any computer that doesn't come off the shelf with sane defaults. Everything else is arcane to them.

We are not those people. I have to remind myself that not everyone likes to build their own systems.

I do have a friend who wants to buy a framework laptop with Fedora on it because that's what they use in the Laboratory he works in but he doesn't want to assemble it himself he just wants it to come like that.

I think we're getting there finally.

I think the gap between what the average Linux user thinks is ease of use and what the average non Linux user thinks is ease of use is probably much larger and many devs seem to understand.

I think it would be beneficial to have a completely idiot proof installer that doesn't ask you about partitions or formatting or basically anything just point it towards a drive and it will set up a default installation.

More GUI based means of doing basic stuff. A casual who wants to access some photos from his laptop does not want to figure out how to manually configure samba shares by editing config files in their terminal based text editor.

I think codecs are a much bigger pain in the ass than is ideal. As I understand that there are legal reasons for this but the first time some casual goes to play a video and gets an error message their first thought may not be "let me search Google and figure out what this error message means" their first thought maybe "Linux sucks and can't play videos".

The permission structure that makes Linux so secure makes it a little annoying for casuals. For example, you actively and intentionally go to the default software store, navigate to the updates tab, update a package you've already installed and clearly want, and do so from the official OS repository... This requires that you enter your password to protect you from what exactly? It's not a big deal it takes one second to type my password, but how would you explain this to a casual in a way that makes sense? Your OS is protecting you from potentially rogue acts of official patches to your default text editor.

I think the folder structures are pretty big challenge for converts. On Windows you can find most of the files associated with any given program in your program files folder. On Mac there's an applications folder. On Linux... it's somewhere, don't worry about it. That's not really a fixable one it just is what it is.

I guess it would be reducing the need of terminal usage as much as possible. That's still the only thing a common user struggles with, in my opinion. The rest is just difference or has nothing to do with Linux.

With Linux gaming is rising currently, most common problem is kernel anti-cheat games and it's not Linux problem, for example. What are devs supposed to do? To develop literal Windows kernel compatibility layer or something? But Linux may do stuff on their end to make cheating difficult to keep game studio's happy but that would also mean to stray away from its philosophy. As a general platform, it would be hard to do this anyway. This would be possible per distro basis. Maybe Linux dev circles are already discussing this, maybe not, I don't know honestly.

People who convince themselves they "just aren't good with computers."

In the early 2000s, it was widely thought that everyone who grew up with them would be reasonably competent with them. We now have 20-30 year olds who are still stumped with basic computing concepts like how to reset a forgotten password. I literally ran into this a couple of months ago: Really? You haven't had to do this a dozen times in your life by now? How did you finish college (this person was highly educated)?

Be preinstalled on laptops/desktops.

everything else is ready unless you use niche software. Most people just use a browser and word or a pdf editor.

note the distro MUST be an immutable up to date kde flatpak using one for normal people, however

A multi-billion dollars marketing budget, anti-competitive practices and confidential agreements, blacklisting hardware vendors if they dare proposing an alternative, and of course a legal department the size of a small city to sue all competition out of existence.

Oh wait that's Microsoft/Google/Apple/Meta/Amazon.

I think it's more users need to realize that an OS that is easier to use in every way is not a more difficult OS to use.

But also, I'm okay gatekeeping Linux, as bringing the masses over just means enshitification and turning it into Windows again. Fuck that.

I think with any alternative to big tech the problem is most people are really unwilling to change their habits and make short term compromises.
A lot of people know on a surface level that big tech is stealing their data etc.
But actually changing their habits goes to far.

Another issue is that its more or less a systemic issue.

To many people aren't even awear of what FOSS even is.
The state of Foss and is a bit complicated where you do have organizations and activists advocating for it but also gigantic corporations that use Foss technology and exploit the free labor that goes into it.

There definitely needs to be more activism for FOSS technology and alternatives to big tech.
And those alternatives should be open to everyone like Linux is.
Of course there are always multiple reasons why something isn't used but I do think it is important to look at a bigger perspective than individual consumer/ in this case users

Adding my voice to the hardware compatibility issue. While most hardware just works, Linux usually lacks the ability to configure the device. Audio interfaces are a good example of this. They work but you can't set the sample rate or enable any custom features on ANY of them.

I believe government regulators should step in and require hardware manufacturers to provide Linux support equal to Windows or Mac. This could be relaxed for low volume or highly specialised devices, but mainstream consumer stuff should be more universal.

I'm reminded of a video I saw of a woman talking about her dating prospects using M&Ms. She poured a bunch on the table as a metaphor for her dating pool, and slid away M&Ms as she ruled the people they represent out. "8 million people in the city. But half are women slides half of the M&Ms away of the remaining 4 million men, 20% are under 25, slides more M&Ms away" until she got to a point where she had one candy left, and then she shattered it with a meat tenderizer and continued sliding pieces of it away.

You can do that for potential adoptees of Linux, because there are a bunch of filters in series you have to pass through before successfully adopting Linux.

8 billion people on the planet.

Subtract the Sentinelese and Amish and North Koreans and everyone else who just outright doesn't have access to computers. Nothing we can really do about them and in some cases it would be unethical to try.

Now subtract out the people who only use a mobile device like a cell phone or tablet, which are locked to their OSes. Android or iOS is as much a part of the hardware as a microwave oven's firmware is to them. Linux on mobile devices (excluding Android) is in a severely rough state, there's basically no hardware and software combo that is ready for daily driving.

Now subtract out the people who do use a PC or other device, that won't ever install an operating system on a computer themselves. You'll get some of these folks by selling computers with Linux installed in stores and such, though I think you'll have to address a few other points later. I think SteamOS is demonstrating this.

Now subtract the people who might install Linux themselves, say PC builders who would have to install an OS anyway, but bounce off the process of choosing a distro and then installing. The big distributors like Canonical and Fedora tend toward marketing wankshit instead of human language. You can't tell their goddamn websites "I just want the normal end-user desktop version with KDE please." Does "Core" mean our main, central product, or the IoT embedded system version? You kind of have to know Fedora calls their Gnome edition "Workstation" and if you want "normal Fedora but with KDE" that's a "Spin." Then you get the Trendy Fork Of The Month, things like Bazzite and Nobara that pretty much are Fedora or Ubuntu with a theme applied, maybe some actual features in the OS, but often just a redone onboarding process, like I think it's Bazzite that offers a configurator on their website that lets you pick your desktop and such. Defuckulating the onboarding process of major distros might allow us to do away with the Trendy Fork Of The Month.

Now subtract the folks who get a Linux machine up and running and then bounce off of the unfamiliar UI. I'm pretty sure this is Gnome's fault more often than not, Gnome is deliberately hostile to both distro maintainers and end users to the point there are now four DEs that are "We can't do this anymore" forks of Gnome: MATE, Cinnamon, Unity and Cosmic. You'd probably see more people stick with Linux if it was less easy to stumble dick first into Gnome.

Now subtract the people who got this far and then said "My CAD/art/music/office/finance/whatever software doesn't run on this." and had to switch back. In a lot of cases, software like that exists in the FOSS ecosystem but it's significantly inferior, like FreeCAD or GIMP. These are often kept in a deliberately shitty state because some opinionated programmer likes how the code they wrote in 2004 looks in their IDE, so open software continues to be unadoptable and people continue to pay subscriptions to the Captain Planet villains in charge of Microsoft, Apple, Google and Adobe.