Okay, I wasn’t sure if I want to respond here, the Deprecation thread, or the Bazaar thread. I don’t think it’s fair to Bazaar’s dev to express what is actually the culmination of my doubts towards Universal Blue there, and this thread’s original post is more forward facing than the Deprecation thread, so I’ll respond here.
I’d like to preface that this… Kinda got more emotional than what I originally planned. But as I think about what I wanted to convey, confusion and skepticism started to dig at “old and recent wounds” as it were. I am very sorry about this, but I’d like to honestly express my doubts here. Regardless, here goes what I originally wrote for the other threads, combined:
I won’t lie, I was and currently still having a knee-jerk, “Oh, great, it’s yet another GTK app that’s going to be forced on me, and watch as Discover gets hidden, and I have to fix that on my builder when I’d rather just use the DE defaults,” reaction towards Bazaar. I would be more excited if it’s something that looks at home on both GNOME and KDE, which would really set it apart from elementary/PopOS’ app store, which is another option as well. Alternatively, if there’s a firm commitment to trying to make it be adopted by Flathub itself- that would be cool and genuinely help enough new users and many Linux users that I can tolerate it.
Mind, it’s not like I can’t see the point made when Bazaar was brought up in that hour-long video. Yes, it uses the “modern design” that a lot of people expect these days (even if a lot of others aren’t big fan of it). Being able to blacklist certain apps would be powerful, if wielded wisely and not in a way that angers people (let’s not forget people still do not like “immutable” OS for “blocking” them from installing packages “the normal way”). Better DE integration (assuming that works on KDE as well) would be great. And of course donation is very important. Speed too, last but not least.
But I’m confused as to why this couldn’t just be done through something that already exists. If the problem is with DE defaults not being satisfactory, I’d be more interested in coordinating a bounty for either good DE defaults or Default Flathub GUI across distro, instead of coming out in favour of yet another shiny but not-yet-mature project. I’m not going to knock on the original dev, for making something cool that is then adopted by a larger project. I’m just questioning how does this make sense for Universal Blue, and its products outside of Bluefin, as well as how this actually fits with what was said in the Deprecation thread:
This outright feels like the opposite of the Project Non-Goals section of the Mission statement. When I read the missions, I felt, “You know what? That’s fair.” I can tolerate the decoupling, deprecation, and reworking of various stuff like Fleek, later Nix as a whole, yafti, and the various existing base images. And I can see how ublue focus on working with upstream, what with bootc and stuff. Ptyxis… Ehh, I can tolerate it, I have a builder anyways, I can restore it back to proper good old Konsole.
But between this, the weird “Ooh shiny”-like behaviour with choosing a Distrobox GUI between BoxBuddy Distroshelf and Kontainer, making an entire new installer- I feel like it’s just a bunch of distractions.
Something that could mitigate it is a real roadmap. I was tolerant of all the recent changes, and even still advocate for people to use Universal Blue as a whole, because I believed in that Mission statement. But lately there’s just so many things taken away, meanwhile random new stuff like Distroshelf and Bazaar are just brought up that is just yet more GTK-ism for KDE user like me, and all the while I just find myself really skeptical as to how this is making this better for everyone- not just Bluefin user, not just Bazzite and Aurora user, and not just Linux users.
Please just give me something rally on. Something that make me feels, “Okay, these are a lot of changes, these are a lot of new things that is pushed to the ecosystem as a whole, that will benefit me and the rest of the ecosystem, and is a major selling point that I can proudly say to other users looking into what Linux ‘distro’ to use.” A blogpost, about all planned changes and why it’ll be much better, or something. Instead of this, “new stuff here, deprecate stuff there, vague statements of intents now and then, oops, deal with it.”
As an example of something that can demonstrate Universal Blue’s capabilities in working with upstream to solve problems for 96% of Linux users (instead of new stuff that’ll benefit the few thousands users of Universal Blue) that I really would like to see is something like being able to solve the WebExtension gridlock, which is a thing that is still causing issue towards people welcoming Flatpak as the new ecosystem.
Unless we’re just going to use Bazaar’s blocking capability to block browser installation too now. Which is an unkind thought from me but it is a thing I thought as I remembered the whole Native Host Messenger saga in the context of Bazaar’s blocking capabilities. At the same time, if Bazaar could redirect and automate setting the “right way, with good user experience, everything working in a way understandable by new users,” a la Zorin’s Windows App Support (which is something that I wish more distro embrace, as it solves a real issue for 96% of new Linux users) then I’m on board with it.
Once again, I apologise for “crashing out” or whatever it is called these days. But I want to pack it all in one reply, so that I can get back to helping new users and promoting ublue as a solution to others, with a clear head and information.
And we need to show that Bluefin knows where it sits in the open source world, which is currently, right where we are today. At this moment in time our mission appears to help make this world a reality. We don’t really have a roadmap because we just find the next thing to do. Bazaar was a side quest that ended up being an important part of the main story.