Try putting it on the speaker’s nose and jiggling.

Most of Canonical’s software, including things like multipass and lxd are GPL or AGPL licensed. Even their corporate website is LGPL’d.

Meanwhile at work I’m dealing with the on-device differences between MBR and GPT partition schemes… 🫩

This was always the case. Main and restricted were guaranteed by Canonical, universe and multiverse fully owned by the community. A bunch of paying customers were unhappy with not getting updates to universe packages, so Canonical made a separate repository that would do that for Ubuntu Pro. Community members with access to the universe repository can still upload fixes there.

It’s not that iOS is unsupported. It’s just that they prohibit browsers from implementing the features my webapp needs.

Snaps are more comparable to nix, really. They can provide system services and even your kernel. Flatpaks and AppImages are only really about distributing desktop apps, but the rest of the system still needs to be provided another way.

Neither Flatpaks nor AppImages can provide those.

Flatpaks are only “competing” with a small portion of what snaps do.

What is a good js framework?

It’s not a Red Hat derivative. It’s upstream of Red Hat.

In a way Fedora is like interim Ubuntu releases, CentOS Stream is like LTS Ubuntu releases, and RHEL is like Ubuntu Pro. So if you want to stay away from a US company, Fedora isn’t a great idea.

You mean like by hiring a significant chunk of Debian maintainers, including the most active apt developer, by having their employees maintain a significant chunk of Debian packages, and by explicitly upstreaming their patches to Debian?

I, for one, welcome our vigesimal BDFL.

You people use meat flapping to communicate?

Given that keeping him around was murdering both Tuvok and Neelix, killing Tuvix was technically -1 murders.

Heh you’re right, I didn’t catch the bitwise and, so I thought you were making a TypeError joke about comparing strings. Fixing the and though (which I did naturally when I typed it into my interpreter to double-check), we get the issue that they are just using a string for the time rather than a time object. “Too early” is also a valid entry that gets us an available office.

Baloo is doing much better now.

Canonical employees have been part of making Wayland since before Mir existed. Mir came about because Canonical wanted to have something that actually worked within a reasonable amount of time. The result? A bunch of places where groups with lots of resources will contribute to Linux (such as automotive systems) use Mir because it was actually practical to use within a reasonable timeframe. A lot of the recent progress in Wayland has come because the people who were holding it up for years finally gave up and listened to the practical concerns from experts who work at places like Canonical and Nvidia.

And snaps still do things that Flatpaks don’t, such as being able to package a kernel and system services. So even if Canonical wanted to switch to flatpak, they couldn’t. Not without making changes that are fundamentally contrary to the design of flatpak.

I didn’t mention the Unity thing because while I don’t agree with your assertions, I do think it’s an example of how someone at Canonical’s hate for KDE has prevented them from making the same good decision at least 3 times.

If you want to pretend it’s Canonical vs. nonprofits, you’re going to have an ugly surprise when you find out which corporation hired most of the people who held up Wayland for so long and has its tentacles so deeply in Flatpaks. (HINT: it’s not SuSE)

Alongside bash? https://www.justice.gov/epstein/files/DataSet 9/EFTA00315784.pdf

Mir came about because the people behind Wayland were fucking around for years without making progress. Now that Wayland has actually matured, Mir is a Wayland compositor.

Snaps predate (and do a whole lot more than) flatpak.

1 out of 3 isn’t great.