A post was merged into an existing topic: Issue with recent install
Thatās a big if, given the patching that Arch is doing is to address issues that have not yet reached an upstream release, e.g.:
I wish people wouldnāt assume things when it comes to āstableā releases, patching, etc. etc. etc.
Are you honestly being serious here? How on Earth is this issue in EnOS, caused by an automated hook in EnOS, a fault of the Arch developers? How on Earth is the lack of skill of EnOS users to be able to reinstall GRUB using a chroot (which is an incredibly common task that people have to do regularly, after BIOS/UEFI updates, Windows updates, ā¦, ā¦) a fault of the Arch developers?
This sort of unrealistic expectation of things never breaking, ever, and that developers needs to make knee-jerk reactions to mitigate issues found by a subset of users (even though those issues are trivially resolved), is why open source developers throw in the towel.
If you want something that doesnāt break (except those times when it does) then use a consumer OS.
Well, yes, because updates in Arch didnāt expose a problem, so how would anyone know until it hits sufficient critical mass that it gets reported?
Guess you missed the part where it affected Arch users too? Which is why it didnāt reflect well on the Arch devs.
So every bug or regression in any piece of software reflects poorly on the Arch devs? Very unrealistic.
Hi @npaladin2000
I found this It's not just Arch having problems with Grub
It seems to be Grub developers not Arch.
Oh i know the Arch devs arenāt the only one to grab pre release code from the grub devs. That doesnāt absolve then from doing it though, nor from from how they handled the situation. Really just not a good situation all around, so i kinda understand the desire ofsome people to find an alternative. Heck two bootloaders isnāt the worst idea, redundancy and all
For something as important as a bootloader, when they take near a week to even let anyone know that thereās a high chance for breakage if your use it the way they tell you to, yes. And afterwards blaming it on other distros and falsely claiming Arch users arenāt impacted. Yes. They also said they donāt have enough testers for this, but released it anyway. Thatās on then too. Sad state of affairs, but thatās what happens with something this serious and significant.
You mean rEFInd? While trying it I noticed it was attempting to boot using Grub which I couldnāt understand!
What is the point in bootloader calling another boot loader?
Or Can we have systemd-boot AND Grub?
Itās one bootloader, not the only bootloader. Arch doesnāt tell you to install GRUB, thatās up to the user to decide on.
No, there was zero chance of breakage if you use it the way the Arch wiki says to.
So - unless software is thoroughly tested then it shouldnāt be released to the repos. Gotcha.
I think you have an incorrect view of what Arch is and how things work, generally. Of course, if youād like to step up and join the Arch releng and/or testing team Iām sure theyād be glad of the extra help.
Yeah i have both installed and have UEFI entries for both. Plus rEFInd can call grub too.
A user should always be prepared for this sort of glitch when using a rolling release distro that tends to take packages from upstream quickly, with minimal patching (and preferrably no patching).
From the Arch wiki page on System Maintenance, the section Upgrading the system:
Make sure to have the Arch install media or another Linux āliveā CD/USB available so you can easily rescue your system if there is a problem after updating.
Oh, and by the way:
Users must be vigilant and take responsibility for maintaining their own system.
Also:
ā¦upgrading packages can raise unexpected problems that could need immediate intervention; therefore, it is discouraged to upgrade a stable system shortly before it is required for carrying out an important task.
not zero because if you may had just installed a second kernel before the update and your local grub files was installed some times ago you would need to have to run grub-mkconfig and it would be the same issue ā¦
Or used whatever method for creating the config file - grub-mkconfig
isnāt the only one.
Youāre also talking about an edge-case where youāre making a change in addition to updating GRUB.
Plus if your integrating grub with btrfs snapshots the way the Arch Wiki tells you to do, youād be impacted.
cause of that i used would
And there are 4 kernels in arch repo so possible it can happen
Yes indeed - there are plenty of edge-cases that would expose the issue (which is how bugs are found, rather than happy-cases).
rEFInd offers you the choice. You can hide the Grub boot or uninstall Grub and then rEFind wonāt offer it.
I use rEFInd but maintain a Grub boot as itās generally, but not always, a distros default bootloader.