Manjaro 2.0 Synopsis This document covers the organizational, technical, management, and other changes we (the Manjaro Team, et al) like to see applied to the Manjaro Project. The goal of this document is to serve as a point of discussion, and ultimately, once a consensus on its contents and written goals has been reached, as a guide for the organizational restructuring of the Manjaro Project. Motivation The Manjaro Project has been declining over the past decade. It managed to sustain a sizabl...
Interesting. As a former Manjaro user (several years ago now), my problems with the distro were more with their approach to package management and the AUR. They withhold packages for the main repositories, but the dependencies for AUR packages will always assume the latest packages, so I would constantly get into these dependency deadlocks where I could not install or could not update certain AUR packages because the necessary dependencies were the incorrect version. I view this as a fundamental technical problem with their approach, and was my main reason for switching away.
Hopefully the new structure/leadership will result in technical changes which fix their issues. Though if I am being honest, the vision of a Manjaro with rolling packages is basically just a reskinned EndeavourOS, so I am not sure what they would need to do for me to recommend this distro to anyone.
The dependency issues seem like that are a flaw in the Arch design. It is the only package manager I’ve seen that requires running the latest available version of packages.
This was exactly the same for me. Every Manjaro install I had broke sooner or later because of these dependency issues. After my 3rd or 4th try, I decided to switch to EndeavorOS which is extremely stable for me and serves me well for a couple of years now.
I’ve used Manjaro and, over time, it’s left me without GRUB and without a graphical interface on several occasions – just as has happened with CachyOS, EndeavourOS, Arcolinux, and others. That’s why I no longer use Arch or Arch-based distributions. I admit that, in my opinion, Manjaro is the best Arch-based distribution, provided you don’t install anything from the AUR repository. The problem is that Pamac and some of Manjaro’s own tools don’t follow Arch’s dependency rules, so that mix of Manjaro’s own repositories and Arch’s original repositories can be a problem.
I just avoid the AUR on Manjaro whenever possible. It still works 99% of the time. The few things I actually need to be bleeding edge I will just try to build from source.
IMO they should have made this the official policy instead of adding optional support for the AUR in pamac.
At the end of the day, the AUR is just a pastebin full of pkgbuild files for people who know what they’re doing. And as a distro aimed more at the average Linux user, rawdogging the AUR probably just shouldn’t be part of the equation.
Could they not have created an AUR mirror and delayed that to be in sync with the main repo’s? It would have solved the AUR ddos that the Arch team got mad about a few times and the out of sync dependencies.
The AUR just hosts pkgbuild files, no source or built packages. The pkgbuild can point to arbitrary external sources that could update separately. Manjaro could have their own AUR that hosts old pkgbuilds, but that wouldn’t be foolproof since the external sources could change. Also, if a pkgbuild was updated for security reasons, now Manjaro is putting users at risk by continuing to serve the old version, and now that’s another problem for them to solve.
Hold up, isn’t that last point just a criticism of delayed updates in general? By that logic, would Manjaro be putting users at security risk by holding back the main packages?
Considering they just hold back packages, but do not do additional testing to release them, yeah, they should not do that.
Arch already has testing repo, normal repo packages on arch are already stable enough