Already used these drivers on previous installation, was 525 iirc (Linux Mint), but also went from 530 to 535 on Arch and it persisted. Thing is problem started the exact day I also flashed BIOS firmware so it’s likely that it could come from there, but trying lts kernel now
It was from a GitHub Gist but idk which exactly it was, there are multiple. Keep in mind some files need to have copy-on-write deactivated (swapfile, VirtualBox disk images). The Arch Wiki mentions when copy-on-write should be turned off for a file
What do you mean with “birds part”? Learned from YouTube Videos, Arch Wiki, and experimenting on bare metal and in Virtualbox. Hardest part for me when installing Arch 1st time was partitioning and bootloaders
You might install an older kernel version from /var/cache/pacman/pkg and then regenerate the initramfs. If not using NVIDIA, it’s very easy to have multiple kernels installed (e. g. linux, linux-lts) to have another option if one kernel causes trouble.
I’d generally recommend having the lts or mainline kernel additionally if you use custom kernels, like zen or self compiled
In the Gentoo wiki it is also mentioned that “While it is true that Btrfs is still considered experimental and is growing in stability, the time when Btrfs will become the default filesystem for Linux systems is getting closer.”. I don’t know how many distros out there use Btrfs by default (never distrohopped), but it seems to become much more widely adopted than zfs.
I wrote this more or less for fun; it is slightly more extensive than the installation guide geared for a more advanced setup. The wiki is mentioned in the article as well and is encouraged to be used too
The Bootloader itself cannot be encrypted afaik, but the Kernel and initrd can reside on a LUKS Volume (GRUB_USE_CRYPTODISK). But, in order to prevent having to input your passphrase twice, you need to use a keyfile, and I have no experience with that, so I have gone another route. I don’t think that a kernel and initrd necessarily need to be encrypted
Do you have GRUB installed into the ESPs fallback path? (esp/EFI/BOOT/BOOTX64.EFI) I haven’t tried grub-install --removable yet, but maybe stuff got confused.
Had the problem only on one machine. Do you have, by chance, a MSI motherboard? Can’t myself think of other causes and having the kernel and initrd on btrfs instead of ext4 can’t be the problem?
Did grub-install /dev/sdX, x86_64-efi and ESP get detected automatically and no errors were reported. Also created grub config. Will try in a few days again. Maybe I really overlooked something or had “bad luck”. Worked fine in a VM.
Could the BIOS firmware have a part in this or is the BIOS firmware irrelevant to the bootloaders functionality?
Maybe Guides / Information could be shown on the lemmy web frontend by default to lower the entry barrier?
Just looked. First 1/2 loads were slow but after that it’s lighting fast! I think by not everyone establishing a Websocket connection and just loading once performance should increase a tad bit.
Try pacman -Syu or just yay, if it doesn’t work after that get new mirrors from here: https://archlinux.org/mirrorlist/, and paste them into /var/pacman.d/mirrorlist. Comment out/in mirrors as needed. If you did the latter, do a full system upgrade (yay) again just in case
Yeah flatpak packages bring their own runtime packages so they’re more independent of the underlying system. I installed the steam flatpak now and it works just fine
When I upgraded lib32-nvidia-utils was already at 535, and the problem itself is still there for me. The probable cause is libcef invalid opcodes (dmesg)
traps: steamwebhelper[1959] trap invalid opcode ip:7f6575bdb794 sp:7fffa9a5d930 error:0 in libcef.so[7f65732ef000+7770000]
I assume that there is something that is O(N), which explains why wait time scales with community size (amount of posts, comments)
The one in the post description? Seems to be a bug, will probably get fixed later.
If you didn’t tell I wouldn’t even have noticed, awesome work!
I got a random “Report created” notification in the bottom left of the screen while viewing this post? Has this happened to anyone?
Only difference in info I can see is that display name ends with :0 instead of :0.0. Depending on your DE, you might fiddle with display settings there. Are you running X natively and not XWayland?