What even is your purpose for existing as a distribution when people can more or less do the same on Fedora? Nobara at the very least has its gimmick of being a Fedora-based Linux distribution focused on gaming. I've yet to really hear about any defining qualities which set Ultramarine Linux aside from the rest other than being yet another RHEL derivative.
Just thought of something interesting worth sharing that you could do to distinguish yourselves from Fedora while continuing to provide the quality of life changes that you are known for @ultramarine. You could rebase yourselves on Fedora Rawhide to become a rolling release in a similar vein to OpenSUSE Tumbleweed while providing the stability and ease of use that Ultramarine Linux is known for.
I believe this would be a very cool and meaningful way to set yourselves apart as a standalone project and not just another Linux distribution riding on the coattails of RHEL as Red Hat likes to think of their clones. In this sense, you would be more akin to Manjaro who are doing their own thing independently from Arch Linux despite being based on the latter.
Thought I would provide some tangible solutions to coincide with my constructive criticism from the other day.
Edit: Sorry for the multiple notifications for those tagged. 😫
Ultramarine Linux aims to be a replacement OS for any computer newbies; we don't just target gamers. You can game on Ultramarine, that's absolutely fine. You can game on any distros too provided that they're good enough. If you can't game on a distro, maybe it kinda sucks.
If you can do the same thing Nobara does on Ultramarine, why bother? If you can run most machines on Ultramarine when compared to Nobara, why Nobara?
Fedora Rawhide destroys the point of Ultramarine. Ultramarine is a distro for everyone. Rawhide is not for everyone: it's very dangerous to run a system on such an unstable repository that rolls things out maybe even faster than Arch does. We can, however, leave that for Project Aoyama instead, which is a very early-stage planned distro led by me and a few other people.
Switching a Fedora/Ultramarine stable system to Rawhide would not provide better stability nor ease of use. I don't see any correlations. That's the first point.
Secondly, relatively speaking, we do not rely on RH as much as most people outside think we do. We have been making so many substitutions in Fedora's infra, including a separate build system, a separate ISO-image generator, a separate installer (available in UM40), etc.
I am not so sure what you mean by "Manjaro who are doing their own thing independently from Arch Linux". Yes, Manjaro has its own AUR helper pamac (iirc), it provides ISOs with desktop environments preinstalled, it has its own distros, but if you really think about it, it's just Ultramarine right now, but Arch Linux. Manjaro provides some sort of rolling release (that get delayed a week for some reason?) so you get to enjoy very new software while using an easy-to-install environment. Ultramarine is supposedly even better, it's harder to break by itself compared to Manjaro.
If you have more questions regarding Ultramarine's philosophy / stability, you can read our FAQ page. Personally however, I welcome any really fancy suggestions to Project Aoyama that might not be suitable for Ultramarine.
N.B.: Project Aoyama is a separate planned experimental distro not really to Ultramarine. Yes, it's developed in Fyra Labs, but there are a lot more outside contributors.
And to clear any confusion on "why Ultramarine":
high stability (unlike Nobara and Manjaro and etc.)
sane defaults (unlike Fedora and other derivatives)
OpenSUSE Tumbleweed which I cited in my previous post is an example of a rolling release done right. Nothing is stopping you from providing a rock-solid and stable experience on a Linux Distribution derived from Fedora Rawhide. How stable a distribution is ultimately comes down to how stable you want to make it. You could, for example, have an atomic filesystem using BTRFS snapshots and that alone would drastically reduce the likelihood of something breaking causing damage to the system and inconvenience from that.
I mean there are probably a million Arch users out there daily driving their Arch install like nothing.
And the only thing stopping us from providing a rock-solid and stable experience on a Linux distro derived from Fedora Rawhide is getting people paying our bills. And userbase of course.
Also we use BTRFS, but somehow dnf transactions don't actually make a btrfs snapshot on Fedora. It does on OpenSUSE though.
OpenSUSE uses something called Snapper which has functionality built-in to provide atomic transactions whenever zypper is used to manage packages. There was an AUR package which aims to replicate this functionality on Arch Linux so it shouldn't be too difficult to either fork the package or create a script to do the same for DNF. Back when I transitioned from Manjaro to Arch Linux proper, I used BTRFS with snapper and pacsnap to provide atomic transactions for pacman which combined with the option to boot from previous snapshots in the Grub menu at startup made for an incredibly stable system.
That's why I personally don't buy it when people say that Arch Linux or that the rolling release model is too unstable or unreliable to use in production. OpenSUSE Tumbleweed manages just fine, so it should be possible for other Linux distributions to follow in their footsteps and achieve similar results without compromising on their mission statement.
@madomado@Ailunarenth I think there’s a serious misunderstanding of what rawhide is and who it’s for. Rawhide is not stable for day to day use and is only similar to openSUSE in the fact that it’s rolling release. Fedora upstream doesn’t care if they break rawhide because it’s not the point of rawhide to be a stable system, it’s for Fedora maintainers and devs, if it breaks they can recover or it’s in a VM so it doesn’t matter.
Additionally, if you’re proposing holding packages back like Manjaro does, that costs lots and lots of dev time, and has proven time and time again to cause more issues than it’s worth.
You’re asking for us to distinguish ourselves with a unique release model, but this model you’ve proposed would require us to give up the extreme stability that is foundational to our distribution and spend dev time maintaining rolling packages instead of developing features (that actually serve our mission)
Switching to Rawhide would be a move purely for the cool factor, and would not match our values at all.
Ultramarine could just not be for you, and that’s fine, there’s tons of fish in the sea :3
Add comment