I don’t think there were any standards for this sort of thing when Flatpak and Snaps first came out, and they arrived at the same time and have pretty big differences. So this doesn’t apply.

Because AI can’t generate 1/10th of the code I write.

When it works I totally use it. Fixing the hallucinations and bad code is usually quicker than writing from scratch myself, especially in a language or system I don’t use regularly.

Well that’s clearly worse… Why even make this comment?

That’s way more than 30 lines or code and also their “We didn’t add anything.” claim seems to be nonsense.