You are wrong. I doubt I will change your mind.

1. There are many, many, many more companies using Linux without giving back than there are for BSD. And not just “using it” either. Practically the entire embedded universe is one giant GPL violation.

2. Linux is not “true” GPL anyway, so it is a poor example for how the GPL impacts success.

3. The companies that build businesses on FreeBSD tend to give back. There are many examples, the biggest being Netflix.

4. The classic example of a company not giving back is Sony and even that is wrong.

5. People choosing a BSD license value different things, rendering your entire premise meaningless for them and your framing of “the problem” inappropriate.

I do not want to get too deep into Sony. But let’s acknowledge that they first tried to ship Linux on PlayStation. They had to stop. Why? Well, it was not because people tried to copy the operating system. It was because people used it to circumvent other protections to copy proprietary games. The problem was not with Sony’s ethics but with those of “the community” and the lack of respect “the community” had for the concept of copyright.

So, Sony switched to a FreeBSD base and they no longer share that code. True.

However, Sony does contribute to BSD. And Sony is a significant contributor to Clang/LLVM and they do share their work freely (even though the license does not require them to). The FreeBSD project benefits from this as Clang is the system compiler. I benefit from this as my Linux distro also uses Clang as the system compiler.

The BSD license is “free software” and provides all “4 freedoms” touted by the FSF. It protects your rights with regards to the code you have and are using. It does not give you guaranteed access to FUTURE code that you do not write. Those future contributors are free to choose their license. You know…freedom.

BSD lags in features, particularly hardware support, because it has fewer users and therefore fewer developers. That is mostly an accident of history and not, in my view, due in any way to the license. Look up the BSD lawsuit that was happening when Linux appeared. If your argument for the popularity of Linux is the GPL, why did Xorg become the dominant window system instead of something GPL based? Why did Rust, Swift, and Zig appear on LLVM instead of GCC?

Anyway, I could write 100 paragraphs and not change your mind. You certainly have not changed mine.

Yes, the big feature seems to be their package manager. But just because an update succeeds does not mean it did not break anything.

They also have their own boot manager and they seem to be fans of Rust, which explains the COSMIC desktop option. They have their own build system.

It is not clear to me that they are doing anything novel beyond that.

They do not have centralized configuration as far as I am aware so they do not go as far as Nix. As a Chimera Linux user, the atomic updates and bespoke build system feel like things I already have.

Overall it sounds like a nice project. But the improvements seem more incremental than revolutionary.

The Fedora Project was created by Red Hat explicitly to be a community distribution so that they could focus on RHEL (Red Hat Enterprise Linux) as an explicitly commercial dostribution.

Before Fedora, there was Red Hat Linux (not RHEL). Red Hat Linux was commercial but inexpensive. I remember it being around $50 but that could be wrong.

Red Hat made installing Linux easy back when it was not. They created the Red Hat Package Manager (RPM). They bundled configuration utilities and a commercial X server back when XFree86 was less great and had far worse hardware support.

As Red Hat Linux became more successful, there was increased tension between the “free software” community and Red Hat’s commercial ambitions.

An individual Red Hat user created something call “Fedora Linux” which was a repository of software for Red Hat Linux that Red Hat did not ship (it was not a full distro). Very soon after this, Red Hat announced the creation of the “Fedora Project” as a collaboration with “Fedora Linux” (the original Fedora Linux project was absorbed into the Fedora Project). The Fedora Project shipped a full Linux distribution called “Fedora Core” which was Red Hat Linux with the third-party commercial software removed. On top of the “Core”, the Fedora Project shipped the packages that Fedora Linux had been shipping. After a few releases, Fedora dropped “Core” from the name and we have Fedora Linux again (but as a full distro now).

Fedora was defined as a purely community distribution with a mandate to ship only Open Source software and to exclude anything commercial. Red Hat provided infrastructure and paid Red Hat employees to staff key positions in the project.

The original Red Hat Linux was discontinued, leaving Fedora as an explicitly community distro and RHEL as an explicitly commercial one. RHEL was slow moving and conservative. Fedora was faster paced and innovative. Tech that would later appear in RHEL would appear in Fedora first, often funded by Red Hat or built by Red Hat engineers. This was before CentOS even existed and it continues to this day.

Red Hat has already “taken over” Fedora. They created it. They still largely run it (staff it). But it is critical to their strategy that Fedora be a fully free software “community” distro. That is the whole reason it exists. So the idea that Red Hat will “take Fedora corporate” or “make it closed source” is completely ignorant of the history of the project.

If Red Hat did not find Fedora useful in its current form, they would simply abandon it. They would stop paying employees to work on it.

In the years since the creation of Fedora, the CentOS project was founded. CentOS was originally a “clone” of RHEL. It was compiled from the same sources as RHEL. It was “downstream” of RHEL. It was not created by Red Hat. It was a problem for Red Hat as RHEL was meant to be an explicitly commercial offering.

Red Hat took over the CentOS project. And they completely changed how it worked. They release “CentOS Stream” as an entirely new distribution. It is “upstream” of RHEL. That is, instead of CentOS being created from RHEL (and being essentially identical to it), RHEL is now created from CentOS. This means that CentOS is actually more of a community distro (for example AlmaLinux participates its evolution) but CentOS is no longer bit-for-bit compatible with RHEL.

I bring up CentOS as I see it as instructive for Fedora. Red Hat wants it to be LESS identical to their corporate offering.

How is it unenforceable?

Real question

Agreed. The need for Flatpak goes way down in a distro with access to the AUR.

I use Flatpak for pgAdmin because the Arch packages are terrible. But it is the only one.

He is doing the same kind of rewrite that the Ladybird founder did recently. They are just reacting differently to how well it is going.

Monoculture is bad for everything.

For one thing, it protects you when your BDFL loses their mind completely.

It IS the early days of GNOME. MATE started with the source code of the last GNOME 2 release.

MATE exists because people loved GNOME 2 and hated GNOME 3.

I am going to go out on a limb and say that they like Debian but dislike systemd.

Most of the fundamental packages in your Linux distribution are primarily written by Red Hat. Do you use Glibc, GCC, gnu utils, systemd, GNOME, podman, pipewire, Wayland, Xorg, or Flatpak for starters?

Red Hat is hardly a free rider in the open source world.

It is also worth noting that Red Hat created the Fedora Project. They created it so they could have RHEL (corporate) and Fedora (community) instead of just Red Hat Linux which they had before.

It always makes me laugh when people worry about Red Hat “taking Fedora corporate”. Fedora was created explicitly to be the community offering and is a key part of the Red Hat strategy. I guess not everybody knows their Linux history.

Many of the Fedora leaders and maintainers are Red Hat employees.

As for US influence, that has always been a thing. US law dominates the thinking. What you really need to worry about is the Linux Foundation.

Arch should have this in a few days to a couple of weeks.

By “shares some of the same codebae”, I assume that you mean both COSMIC and Niri use the Smithay Wayland support library? Or that both Niri and GNOME use xdg-desktop-portal-gnome?

Interesting that you are using PaperWM but recommending Niri.

Have you tried COSMIC yet? Maybe PopOS is worth a shot.

Some packages are a bit old at the moment but they have a release coming in April / May that will bring them right up to date.

Perhaps LMDE (Linux Mint Debian Edition) is worth a look as well.

Both options are similar in that they take a very stable distro base and layer on a quite up-to-date desktop.

They also feature clear direction and a predicable release schedule.

Mint or Fedora require no more command line than Windows does.

Probably not a universal answer as you are optimizing for different things.

I will say that EndeavourOS is essentially vanilla Arch once installed. If you really love configuring everything yourself, vanilla Arch is what you are looking for. If you like Arch but just want to fire up a system with sensible defaults, EndeavourOS adds a lot of value without corrupting the purity of the base system.

So, my vote is for EndeavourOS.

Cachy adds the most additional functionality but also changes the base system the most. If you have a T2 MacBook, this is the best option for sure.

I would avoid Manjaro.

Garuda has fans. A bit much for me.

I am not sure why the downvotes.

Redox includes a C library. You should be able to use whatever programming language you want. It does not have to be Rust.

Redox is still adding bits needed for easy porting of more software.

The window system is its own thing at the moment. But they are adding Wayland compatibility.

It will be interesting to see what Redox looks like by the end of 2026. You may get many of your wishes.

You summarized my position very well. Thank you.

I realize I oversimplified a complex set of moves and “shared source” is its own can of worms. My post was already too long.

But my core point is that the code (as Valkey) remained available and remains available under the same free software license that it has always been available under.

The only consequence of what Redis did was that they stopped giving away their “new” code to service providers like Amazon. Even Amazon can continue to use what was there before. And the community can continue to collaborate on the same code base that they were collaborating on before. The licence Redis chooses for its “new” code is largely irrelevant.

We talk about permissive licenses like they represent some massive risk. I just do not see it that way. And they have many advantages including often attracting more corporate participation (more free code for me).

I am a very happy user of Clang/LLVM. It is the product of collaboration between Google, Apple, Sony, Microsoft, academia, and other nerds. I am very happy we have licenses that encourage companies to create quality software for me to use.

I am sure Redis chose BSD to begin with in case they ever had to make a move like they did. If the only option was GPL, they may never have released it as Open Source to begin with. Again, I am glad they did.