Snap came BEFORE flatpak. Flatpak was the "new competing standard" in that situation.
And Gnome shell, quite frankly, sucked. IMO it still sucks, but back then it sucked WAY worse.
At the end of the day the scorecard reads:
- Mir: failed
- Unity: failed
- Snap: mostly failing
Meanwhile RedHat takes over stuff that doesn't work, makes it work a bit better, and pushes it on the whole ecosystem as "the" solution. And they win, and win, and win.
As much as I hate Snap and remove it from my Kubuntu systems, I don't see where it is failing. I frankly see a lot more non-linux-focused vendors support to snap than flatpak. Could you expand on that point?
It’s not a question of which one will succeed between snap and flatpak, it’s whether the ecosystem really needs either one of those.
unity failed because they abadoned it, but it was way better than wayland+gnome. the problem was that it was based on gnome2 and had mir under its belt, so it would've been really really hard to somehow upgrade it
Unity didn't fail: ongoing development on it was cancelled because there was no way to successfully monetize it. It was, and remains, one of the most successful desktops out there.
If your definition of community is "people who develop desktop environments for open source software" then you're already limiting the size of your community to a few dozen or so individuals, and we had a few dozen contributors to Unity so I'm not convinced of the strength of this argument.
If your definition of community is "people who don't use Unity" then of course "everybody knows" that's trivially true. Some people also know it's a tautology.
What exactly is it that you think Wayland couldn't do and why was it necessary to invent an incompatible application interface to achieve that?
> Snap came BEFORE flatpak. Flatpak was the "new competing standard" in that situation.
And AppImage came before snap.
From the mir technical architect (found on askubuntu): https://samohtv.wordpress.com/2013/03/04/mir-an-outpost-envi...
> And AppImage came before snap.
Exactly right! And if the ONLY goal was compatibility, we all should be using appimage over snap. But snap was and is trying to promise more in terms of end-user security and transactional updates from the vendor. So there is a legitimate reason to make something new.
In general, I personally prefer the way Debian works (Debian the Project - not the Distro). It has a board of elected developers governing the project. I would prefer that over somewhat opaque functioning inside a company (Canonical).
To cite as an example, here's how they decided on the question of init systems [1].
I even tried to install Debian while I'm still not really used to Linux, but the graphics card immediately crapped itself on boot, so it will have to wait...
Nvidia was the least terrible solution about 10 years ago (I have PTSD from installing binary blobs and editing Xorg.conf to make it work.) While others have improved tremendously and you don't have to do anything to get full 2D and 3D acceleration (just boot the system), the Nvidia experience™ hasn't changed much since then.
I'm starting to learn too much about apt to try and prevent things from reinstalling themselves.
I haven't trusted Canonical since I noticed their pattern of creating competing alternatives to new Linux standards instead of helping them (Mir & Wayland, Snap & Flatpack, Unity & Gnome 3). It'd be one thing if they were bringing better ideas and long-term support to their alternatives, but they just seem to be half-baked copies. I appreciate all they've done for the Linux ecosystem, but I'll stick with my Debian.