Do I need to be concerned that any EndeavourOS related software will break with system upgrades/updates or is it just the general system maintenance I would have to do on regular Arch?
I was watching a video by Eposvox where he switches his gaming machine over to Garuda Linux and has a great time for the first few weeks, then a system update breaks everything.
I am aware of the notion that it is the users responsibility to ensure that updates won’t cause breakages, but what Eposvox mentions in his video and what was brought up frequently in the comments was how Arch based distro’s are more susceptible to bigger system issues due to updates compared to Vanilla Arch. How true is this for EndeavourOS?
I really love Endeavour and don’t particularly want to switch over to a base Arch install, I’m just wondering if I should be more careful with updates and upgrades.
EndeavourOS has very few distro-specific packages. Almost everything comes from Arch. The chance of a EOS-specific package breaking your install is relatively low.
Garuda and EOS have very different philosophies and approaches to package management.
I can’t speak for everyone, and I am not an EndeavourOS development team member, but I have been using EndeavourOS for several months on multiple different systems and I have not encountered a single packaging problem.
While no system will guarantee perfection I will assertively state that this has been the most stable Arch based distribution I have used, including Arch Linux and Manjaro, with brief visits to a few others and numerous Linux distributions. EndeavourOS compares very favorably with any of these.
Thanks, I had a feeling that this would be the case, and I know that Garuda is very bloated with their custom software and chaotic AUR I think it is? But it’s good to have confirmation.
I would say this. When you install Garuda, there are probably some packages you won’t need. It is cleaner to build from ground up to install things you do need, rather than installing a distro then removing the things you don’t need.
In case of breakage from updates, it could be few days, could be months. Depends, the more packages, the more updates and more chance. But you have to enjoy the learning and process of rolling back snapshots, researching fixes then you will find it is time well spent.
Give it a go and after few months trial you actually finding it too inconvenient then just switch.
This can have many causes (e.g. users expecting a “gaming OS” to be a zero-interaction solution), and Garuda includes a snapshot/rollback tool for this reason.
Also be aware that YouTubers tend to produce content that generates views, so will generally focus on something “trending” while presenting hyperbole and strong emotions (love and hate make for great “community engagement” and therefore algorithm attention and monetisation). For example: “Is EnOS the best OS ever?!” vs. “Why I’m leaving EnOS for good.”
This all said, EnOS follows Arch much more closely, so if you’re happy with the Arch maintenance approach then EnOS will work for you.
Including an additional repo in pacman.conf is bloat now? Wowsers.
Unless you’re using an old Nvidya card, changes of something major breaking on update are very slim, next to none. To knock on wood, I’ve never had anything break on update, on any Arch-based distro I’ve ever used. At least not something that I either couldn’t fix myself in a matter of minutes, or something that wasn’t fixed on the next update.
The fears of system breaking on Arch are mostly unfounded, in my opinion. Maybe I was just lucky.
Not everyone supplies such useful and focused repos to add as you do! Of course, I speak from a biased position, as a supplier for a chaotic-aur mirror! I think it would be fair to say that chaotic-aur occasionally (but usually without harm) lives up to its name
I don’t like the concept of Chaotic AUR. I don’t think building of packages should be automatised. I could be wrong, but that doesn’t sound particularly safe to me.
I admit, I don’t know much about it. For example, I don’t know who, if anyone, checks the packages, the PKGBUILD file, the sources… It seems to me that it would be quite simple to distribute malicious software this way.
I don’t think it is quite the way you suspect - there is automation, but not blind. Quite a few of the packages are actually maintained in the AUR by the chaotic team, and they keep a pretty good eye on it. Another subset of the packages is those required/belonging to Garuda Linux - many of those are created by people who also help with the repo…
I completely agree about the ‘general’ advice, though - PKGBUILDs are there to be examined (for your own good!).
Interestingly, only a couple of my 14 active Arch-based builds have default access to chaotic-AUR - a lot of the content is more specialized than I need (gaming goodies, nvidia stuff, custom kernels etc etc) - and it isn’t worth the effort for only 1 or 2 things…
A PKGBUILD is only a kind of active Makefile. It can of course do some bad things,
but a much better place to hide malware is the real (often huge) source code.
This. As I am lazy, I usually just try to pinpoint the package that has broken the system, downgrade it back to the previous version that worked and add it to ignore for a while, waiting for the issue to fix itself (i.e. hardworking developers/maintainers figure out and fix the issue, for which I am grateful). Every once in a while I then check to see if there is a new version and if it fixed the issue.
I’ve managed to get over most issues (which I can count on the fingers of both hands) I have encountered due to system update in the last three years by using this method.
A bit of advice from a newbie to Linux as of late 2020. Things are going to break, in every distro there is always a chance. With every Kernel update there is a chance too. I have distro-hopped like mad since 2020. I’ve slowed down now, but enjoyed the ride as everything is a learning experience. I’ve been on several dozens distros and used almost all the desktop environments out there.
I did it all on bare metal, never doing it on a virtual machine because I wanted the full experience and see where things went right and wrong with my daily activities, personal or work related. Sure, lots of setup, wiping, setup etc.
Practice makes perfect and I’ve had a lot of practice redoing it over and over. It isn’t a game-killer. It doesn’t take much time to reinstall, setup, patch. Way faster than a Windows install or MacOS. As long as you don’t put any personal files you don’t want to lose on your install drive you should be safe killing and restarting if an issue happens.
Make sure you have a backup machine to get another ISO to flash to a thumb drive if your main machine dies etc. You’ll survive. All my music, movies, documents, photos and anything else I don’t want to loose is on another drive separate from my OS drive and NOT in my Home folder. Oh and don’t do an update if you have a project due
My personal advice is you are generally better off not doing this. Instead of reinstalling when something isn’t right, fix it. While it could be a bit more work in the short-term, this experience is much more valuable in the long-term.
I can’t remember the last time I had to re-install and I do some crazy stuff that is much more likely to break a distro than most people would. On an Arch-based distro, you shouldn’t need to re-install unless you do something overwhelming silly. There is almost nothing that can happen during an update that can’t be fixed fairly quickly.
Absolutely. Reinstalling is only a solution when everything else fails, and you basically give up. It is almost never necessary, and is almost always a missed opportunity to learn by fixing stuff yourself.
Even using something like Timeshift to make snapshots you can restore every time you break something will slow down your learning. However, it might give you confidence to break your system more often…
And @dalto will probably disagree with me on this, because he is using every distro ever known to man (at the same time, since his brain is parallelised), but I think trying out many different distros is not very productive. It’s good to satisfy your curiosity by trying out different distros, but this in itself won’t make you proficient at it, because you’ll end up spending a lot of time learning stuff that is not very important in the big picture: things that are peculiarities of different distros.
If you are at the very beginning of your Linux journey than switching distros rapidly probably will slow your learning curve. On the other hand, if you are at the beginning, how do figure out what you like without trying multiple distros? It is something of a catch-22.
One thing I would recommend someone just starting out to avoid, is switching distros when you hit a problem. You will always hit a problem at some point, switching distros when this happens is completely counter-productive. All distros are ultimately Linux. Unless you are using something very obscure or limited, the problem can typically be solved on the distro you are already on.
Once you have a good base of knowledge and are past the beginner phase, trying different distros can actually help with your proficiency. Learning how different distros do things often gives you a greater understand of how things work in-general. Of course, this assumes the distros are sufficiently different. Switching between a whole bunch of Arch-based distros is probably pointless. However, switching from distros based a mix of Arch, Debian, Fedora, etc can definitely broaden your horizons.
I think it would be fair to say that chaotic-aur occasionally (but usually without harm) lives up to its name 