Home Creators Posts Import Register Favorites Logout

Downloads

Content

Hey everyone!

As Linux seems geared towards being able to run all Windows, and macOS apps that matter, I decided to talk about what else is needed for Linux to keep growing, and get a reasonable enough market share: basically, what other work is needed to keep gaining new users and let them escape Windows or macOS ?

I hope you'll enjoy listening to this one, and have a great week!

Comments

Eric Flores

What's left? - "Custom Build Software" - For example I develop point-of-Sales software for small companies - payroll, business ledgers. That's where the money is :D Scanning software already done by me for my personal use I have never released it. On printing Linux does just about any printer you can trow at it with some exceptions. However, the majority OEM do not support our platform, this is where the problem is. I believe that's going to change soon.

UsernamesAreHard

What's missing for desktop Linux to grow is to put desktop Linux in the hands of people, without buyers having to go out of their way to get linux devices. Once enough people own Linux PCs then software vendors will follow, they'll go where money is to be made. I think that we can't count on off the shelf hardware retailers because selling Linux PCs just doesn't make them as much money as selling windows or MacOS, they can't sell BS subscriptions like antivirus, office suites or useless services. And honestly, if desktop Linux turns into such a corporate/bloatware mess thing, maybe it's better if desktop Linux doesn't get too popular. One last thing that's very important, is to have a proper package distribution system so that software vendors can actually target desktop Linux as a whole and not one or two distros. It's 2026 and shipping apps on Linux is still a unfathomable mess. We have debs, rpms, appimages, flatpaks, snaps just to name the most well knows ones. It's insane and just not appealing to release software on Linux. Flatpaks were probably the most promising format of them all to ship software widely and it's been stagnating way too much. Big software vendors just aren't interested by flatpaks and the mandatory sandbox is to blame for a lot of it, Flathub's vetting process being probably the second stopper for wider adoption. Flatpaks should have been unsandboxed by default to attract developers and software vendors and to let them build their apps for Linux. The current sandbox offers little benefits for security: it's set up by the flatpak's developer, it doesn't prompt user when an app requires access and when it's too restrictive users just loosen up the sandbox anyway, with the convoluted mess that are Flatpak permissions.

Nick - The Linux Experiment

yeah, hardware is the major stop, I think. Flatpaks should absolutely have offered an unsandboxed option so it really could become THE default, and then devs would have adjusted to the sandbox in time. They tried doing too much all at once, and the transition is complex. It still works, and it's still a massive packaging format in the entire Linux ecosystem, but it could be bigger.

tkf

Hey Nick! I really enjoyed the thought experiment, and I mostly agree on the "next goals" (hardware support, desktop features...), but I had 2 caveats in mind the entire time: 1. Phones: the mobile market is enormous, maybe even bigger in terms of total user time than the desktop, arguably more locked down than the proprietary desktop OSes that Linux has been competing with, and none of those trends look like they're slowing down anytime soon. Yet Linux phones still suck compared to even custom-ROM Androids, and the app compatibility battle is going to be at least as hard as the Windows one... We may not be so close to "running every app that exists", especially with some services operating exclusively (or almost) through a mobile app (the neobanks, fitness apps, messengers, concert tickets apps, flight booking apps...) 2. This kind of follows my previous argument, but in parallel of the incredible progress Linux has made in the last decade (rise of Linux gaming, media coverage, AI, even some EU regulations...), the corporations haven't been twiddling their thumbs either. Stuff that would have sounded completely outlandish 10-20 years ago (remote hardware attestation with TPMs built into SoCs, Android moving towards iOS-level app vetting, even the kinds of anti-cheat we see today) is now normalised or on the verge of enforcement, and I'm really worried that the recent wave of "vibe-liberating" of apps that *actually threatens their revenue* is going to trigger absolutely massive amounts of money being spent for countermeasures, probably with the same AI force-multiplier that OSS is currently using. If every major app (or even some of them, slowly) becomes cloud-first, with Cloudflare-style browser trust checks, or cutting-edge obfuscation, or CPU-level code decryption, the Linux community is gonna need a lot of tokens to compete, and we already know who has the deepest pockets. The recent open-source news make me optimistic, but I don't think AI can simply destroy proprietary lock-in either. (Edit: I found the blog post that this reminded me, it's a long read but probably a better explanation - https://blog.giovanh.com/blog/2025/10/14/a-hack-is-not-enough/ )