I recently came across openSUSE again and decided to give it a try this time. I am daily driving Fedora 40 right now and before coming across openSUSE I wanted to switch to Fedora Kinoite or uBlue Aurora (i.e., immutable / atomic). That’s why MicroOS piqued my interest but I had a hard time find information if MicroOS is suitable for daily driving as a atomic desktop or mainly used for a container host on a server.
If someone has personal experience with openSUSE or could link me to a nice write up comparing the two I would be very thankful!
Edit:
In the MicroOS portal it is described like this:
Rolling Release: Every new openSUSE Tumbleweed snapshot also automatically produces a new openSUSE MicroOS release.
I am a long-time Tumbleweed user. It’s the most stable rolling release distro I’ve tried, so if you want that latest software, it’s a great choice. I’ve not tried MicroOS yet, so I can’t comment on that.
Thank you.
In the MicroOS portal it is described like this:
Rolling Release: Every new openSUSE Tumbleweed snapshot also automatically produces a new openSUSE MicroOS release.
So it should get the latest software pretty fast too, right?
I imagine so, buy have no first hand experience.
I can’t speak to MicroOS but I have been running Tumbleweed for about a month. normally I run arch.l, but wanted to try something new for a change, and I was interested in trying out a full DE as I typically run sway.
I’ve been extremely impressed with KDE; I assume you feel the same if you’re looking at Kinoite, but feels worth saying out loud for other readers.
Tumbleweed, for an Arch user, is fine. it installed fine, was reasonably sane out of the box (although defaults to X11, not Wayland) and it’s been perfectly stable for the month I’ve run it. Doing development on it is very easy, and it comes with a non-root docker setup script out of the box which is nice, and I’ve had no issue building software on it. YaST is powerful but has an awful UI.
However: it has the same problem as Ubuntu for me, which is that if you want software from outside the repos you have to trust other repositories and trust their keys, and they often want to replace packages, and finding out if they are built safely can be quite challenging. compare this to Arch, where you can easily read a PKGBUILD and they almost always download sources direct from the developer/vendor, and they very rarely replace other packages. So I find it hard to trust this system’s integrity over time; where are my packages coming from? So in the end I’ll probably go back to Arch, or maybe try out Endeavour, but if this doesn’t concern you then I think Tumbleweed is a capable distro that’s easy to get up and running.
Yes I already use the Fedora KDE Spin right now, it’s awesome!
I didn’t know that it uses X11 because Fedora uses Wayland already for a few major releases.
I think I found a solution for your problem recently. Are you familiar with distrobox? AFAIK you can use it on top of your OS, in this case Tumbleweed, and install another OS in a container, like Arch, and then export the programs installed from AUR or whatever to your host OS.
But nonetheless thank you, I think I should just try it out in a virtual machine or something.
I’ll check that out, thanks for the recommendation. as for it defaulting to X11, it’s no issue because the Wayland session is also available and has been absolutely solid for me, I was just surprised that it wasn’t the preferred session by the distro.
What’s the advantage of using wayland at the moment?
I mean that’s a fair question, because I feel like mostly the advantages are, hm, not “theoretical” because it’s an actual advantage, but not something you’ll really encounter day-to-day. better security for example. but generally who cares because if I interact with something malicious I’m probably owned anyway.
originally I was interested in it because of fractional scaling, but I think that works in X11 for the most part now?
at this point it’s mostly about using the bleeding edge stuff so I can help find problems. I do find that when it works it works very well, and the experience of using a Wayland desktop is less wonky: fewer weird rendering glitches when dealing with multiple monitors, connecting and disconnecting my laptop from a dock, etc. I find this works better with Wayland, but I wouldn’t say “so much better that you must move to it today” if you’re happy with what you have.
similarly full-system stability has been better, and I have fewer crashes that take down everything, I feel. it’s perhaps subjective though: I’ve been running it for so many years maybe all I’m experiencing is that the software I run has become better in general.
so: I don’t think it’s a night-and-day life-changing experience or anything, but it does feel modern and stable, and it’s definitely where things are heading so why not get used to it now, and help to improve it, is my thinking.
Agreed. And I do understand wayland is the future, having done studies around X11 a bit. The problem for long time users like me is that there are still expert apps and use cases that aren’t covered by wayland, at least for now. And because the current benefits of wayland are not obvious they will complain if their distros transition to wayland too soon.
yeah with the exception of krita (which runs fine on xwayland, even with a tablet) I’ve been able to run 100% Wayland, with sway for work and KDE for home, but my needs aren’t too wild. I’m sure a lot of users feel like the rug was unnecessarily pulled out from under them; change that feels like a regression even for very good reason will almost never feel like reason enough if it’s your shit that gets worse, definitely.
still, I think you’ve got to get people using the thing if you want the thing to get better. probably more casual users didn’t even notice when gnome moved over, for example. but probably even the most casual user ran into some problem, and that’s a bummer.
out of curiosity what use cases/software has stopped you from running Wayland? I do miss the magic of tunneling an X session over SSH, that felt like dang magic in the early 2000s.
Remote desktops. I think the main complaints are the performance. To me the issue is that with x11vnc you could remote into an existing display, even the login screen. Recent Gnome finally seems to have it for wayland, but afaik KDE still doesn’t have it.
(FYI: either something piques your interest, or your interest peaks at some point)
MicroOS is designed as a server OS first and foremost, but I have read some anecdotes of people using it just fine on the desktop.
You might want to look into OpenSUSE Aeon or Kalpa instead, which are immutable editions designed for the desktop, running GNOME and KDE respectively. Kalpa is in alpha (almost rhymed) but Aeon is in a more mature state.
Thank you very much! I read about those. Maybe it’s time to try out GNOME again, I don’t want to use a too early version of an OS. Altough I fell in love with KDE, especially KRunner!
