And how do you quantify their reduced blame for hiring community members already? As I’ve already pointed out, Canonical has many Debian developers and maintainers on their payroll. While we’re unlikely to ever get real numbers for it, if it turned out that Canonical had a bigger portion of their payrolls devoted to ensuring that community developers got paid than the other companies mentioned, wouldn’t that say that they’re even less to blame?

That’s kind of a non sequitur. Canonical hires a lot of community members to maintain stuff for the community. They also have roughly 1000 employees according to Wikipedia. SUSE also depends on things like xz and has twice as many employees. Red Hat has 19,000 employees. Google depends on xz and has over 180,000 employees.

So if you’re blaming Canonical for not hiring the maintainers of under recognised community projects that don’t have corporate backing, then surely SUSE gets twice the blame, Red Hat gets 19 times the blame and Google gets 180 times the blame? (Not to mention Amazon, Meta, NVIDIA, etc.)

You said Snap is a similar but alternative to Flatpak, implying that it was developed in response to Flatpak, which simply isn’t the case.

Snap predates Flatpak, and it’s clearly a big money maker for Canonical with their commercial customers who want things like confined but upgradable services in an airgapped environment. By the time Flatpak was making enough headway to be considered feasible to use, snaps were already pretty widely used and had several fairly big names like JetBrains, ROS and CircleCI publishing on snapcraft.io.

Flatpak cannot and was never intended to do all the things snap can, such as setting up system services or distributing kernels. So even if the assertion that snaps for desktop apps were a response to Flatpak were true (it’s not), it doesn’t make sense for Canonical to stop developing snap regardless, as desktop apps are only a tiny part of what snaps do.

Canonical’s initial hiring strategy was “hey, you maintain Debian packages. Wanna get paid for that?”

They still employ quite a few Debian maintainers, and I don’t think it’s at all a stretch to say that Debian wouldn’t be as good as it is today if Canonical weren’t paying a bunch of people in part to do Debian develops. Their employee roll includes one of the developers of apt, amongst other people.

Snaps predate flatpaks (though not by very long - months I think, but not years).

Upstart predated systemd by quite a while. In fact, RHEL 6 used upstart.

If anything, systemd is an example of Red Hat NIHing upstart.

It uses gnome. That’s why I use Kubuntu instead.

Other people have issues with snap packages, however I’m quite the opposite and actually tend to prefer snaps over other means of getting apps.

They officially publish the snap, the flatpak and a deb in an apt repo.

Really? When did they make the switch?

The initial comment said that they didn’t like that Ubuntu is owned by a company.

Switzerland and Norway are not part of the EU either.

Tellico

This (well, with alt) has been a standard thing in x11 for decades. KDE kept it (with meta) when it added a Wayland backend to kwin.

The first desktop that I used on Linux was GNOME, probably either 2.0 or 2.2. It was a bit clunky, but it was fine. I distro hopped for a while and discovered Mandrake 9 and thought the desktop was great. This was when I discovered desktop environments. I hopped over to Fedora Core when it was first released and was unhappy with the desktop again.

So I started desktop hopping on Fedora. I tried XFCE, Fluxbox, Openbox, and several others. They were cool, and the KDE experience on Fedora Core 1 was not great. At some point I switched to Gentoo and used the KDE experience there. When Ubuntu came around, I found that while the install experience was good, the desktop was kinda clunky. I ended up sticking with Gentoo. When Kubuntu 5.04 came out, though, I switched over. And I’ve been using some combination of Kubuntu and KDE Neon ever since.

If GNOME had been my only option, I probably would have gone back to Windows. Initially because I found it clunky (and tbh kinda ugly), but more recently because every time I’ve used GNOME in the last decade or so, it feels like it’s lost features I used heavily. Meanwhile KDE has taken a different approach to configurability of trying to cut down configuration options by figuring out what a better option that everyone can agree on looks like. It’s still very configurable, but it has nowhere near as many knobs as it had in the KDE 3.5 days. You know what, though? I cannot think of a single lost configuration option in Plasma that I miss.

So I am strongly in the KDE Korner between these two, and much more weakly favour KDE Plasma vs. other desktops.

Personally I have no issues with a corporate backed distro. My point is that if someone doesn’t want a corporate owned distro, PopOS doesn’t fit the bill.

Not sure why you specify binary-based OS’s. Following Gentoo’s upgrade guide also gets you potentially whatever they want on your systemp

Yes! In fact, Chromium was originally a fork of WebKit, as WebKit was a fork of KHTML. In both cases the codebases have diverged quite significantly though.

The core of Safari (WebKit) is open source. If it weren’t they’d be violating the GPL license of KHTML.

Kubuntu, because it’s the most solid distro I’ve used that meets my needs.

It may surprise you to see how many Debian contributors are doing so on Canonical’s payroll.