January 2026 - Arch-general - lists.archlinux.org

The NVIDIA 590 driver notice
by Mark E. Mallett 17 Feb '26

17 Feb '26

Hi, There was a notice on arch-announce and on archlinux.org/news ( https://archlinux.org/news/nvidia-590-driver-drops-pascal-support-main-pack… ) that confuses me. "With the update to driver version 590, the NVIDIA driver no longer supports Pascal (GTX 10xx) GPUs or older" etc. I don't really know if that applies to me, given the brand/model labels and all. (some detail about my hardware is offered at the end of this message.) And even if it does, the notice is (to my dull mind) unclear. It says: Updating the NVIDIA packages on systems with Pascal, Maxwell, or older cards will fail to load the driver, which may result in a broken graphical environment. That would be bad and makes me hesitent to experiment. Part of what confuses me is this: Users with GTX 10xx series and older cards must switch to the legacy proprietary branch to maintain support I'm using the proprietary nvidia driver already, so what does that statement mean to me? (I'm not running nouveau or other) So I have questions - the one above ("does this mean me?") - the one about already using the proprietary driver - can I simply not upgrade the nvidia driver, keep using the installed one? - If the system breaks, what is my recourse? - Is "Install nvidia-580xx-dkms from the AUR" (as it says) appropriate for me? I know this sounds like a lot of foggy nonunderstanding, and I guess that's because that's what it is. The relevant system has a very old NVIDIA card. I haven't seen any reactions to this, maybe I'm the only one left ;) Possibly relevant stuff from hwinfo: 22: PCI 100.0: 0300 VGA compatible controller (VGA) [Created at pci.395] Unique ID: VCu0.4tATP2A+FnB Parent ID: 8otl.w6ofvOouCK2 SysFS ID: /devices/pci0000:00/0000:00:04.0/0000:01:00.0 SysFS BusID: 0000:01:00.0 Hardware Class: graphics card Model: "nVidia GM206 [GeForce GTX 960]" Vendor: pci 0x10de "nVidia Corporation" Device: pci 0x1401 "GM206 [GeForce GTX 960]" SubVendor: pci 0x1462 "Micro-Star International Co., Ltd. [MSI]" SubDevice: pci 0x3201 Revision: 0xa1 Driver: "nvidia" Driver Modules: "nvidia" Memory Range: 0xfd000000-0xfdffffff (rw,non-prefetchable) Memory Range: 0xd0000000-0xdfffffff (ro,non-prefetchable) Memory Range: 0xce000000-0xcfffffff (ro,non-prefetchable) I/O Ports: 0xcc00-0xcc7f (rw) Memory Range: 0x000c0000-0x000dffff (rw,non-prefetchable,disabled) IRQ: 27 (1392224 events) Module Alias: "pci:v000010DEd00001401sv00001462sd00003201bc03sc00i00" Driver Info #0: Driver Status: nouveau is not active Driver Activation Cmd: "modprobe nouveau" Driver Info #1: Driver Status: nvidia_drm is active Driver Activation Cmd: "modprobe nvidia_drm" Driver Info #2: Driver Status: nvidia is active Driver Activation Cmd: "modprobe nvidia" Config Status: cfg=new, avail=yes, need=no, active=unknown Attached to: #20 (PCI bridge) Thanks for any attention. -mm- (long time no write)

10 15

Arch Linux Wiki responds 429 based on macOS User-Agent
by Joseph Chris 28 Jan '26

28 Jan '26

Hi, While browsing the Arch Linux Wiki (https://wiki.archlinux.org) from macOS, I noticed that the server consistently responds with HTTP 429 (Too Many Requests). After some testing, it appears this behavior is triggered by the User-Agent, and is not related to IP address, request rate, or cookies. The following User-Agent from my browser can reproduce the behavior reliably: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.0.0 Safari/537.36 More specifically, the substring "Intel Mac OS X 10_15_7" is sufficient to trigger the block. $curl -vLo /dev/null https://wiki.archlinux.org/ -H "user-agent: Intel Mac OS X 10_15_7" * IPv6: 2604:cac0:a104:d::4 * IPv4: 209.126.35.81 < HTTP/2429 < server: nginx < date: Wed, 28Jan 202606:28:43GMT < content-type: text/html < content-length: 162 < strict-transport-security: max-age=31536000; includeSubdomains; preload Does anyone have any idea about this? Best regards, Joseph C.

1 0

ex-vi-compat: neovim
by Edmund Lodewijks 27 Jan '26

27 Jan '26

Greetings! Recently the package `ex-vi-compat` was introduced, which conflicts with, and replaces `vi`. This PKGBUILD installs `vim` onto the system, and symlinks a bunch of commands to `vim`. However, I use `neovim`, and it seems a bit superfluous to me to install `vim`, when `neovim` can do all that `vim` can do in this regard. Is there any way to perhaps check if either `vim` or `neovim` is already installed on the system, and then use either one of them in the symlink? Something like: ```sh package() { local cmd local vimcmd # Determine which editor to use if command -v nvim &>/dev/null; then vimcmd="nvim" elif command -v vim &>/dev/null; then vimcmd="vim" else error "Neither neovim nor vim found" return 1 fi install -vDm 755 vi.sh "$pkgdir/usr/bin/vi" for cmd in {edit,ex,vedit,view} ; do ln -sv "$vimcmd" "$pkgdir/usr/bin/$cmd" done ``` Alternatively, perhaps a few changes in `vi.sh` about which editor to invoke? Kind regards, - Edmund -- Edmund Lodewijks <edmund(a)proteamail.com> TZ: UCT+2 / GMT+2

3 4

lxpanel weather plugin icons are missing
by Ralf Mardorf 27 Jan '26

27 Jan '26

Hi, for several hours now, the weather icons have been missing from the extra/lxpanel 0.11.1-4 weather plugin on my computer. The temperature is displayed as usual. Is anyone else affected besides me? I believe the icons are transferred together with the data from OpenWeather, but I may be mistaken. Regards, Ralf

1 1

Welsh / Cymraeg maintenance offer
by fin-w＠tutanota.com 27 Jan '26

27 Jan '26

Hello, I work with the organisation (Bangor University) responsible for creating the Welsh (Cymraeg) Hunspell data (.dic and .aff files). I am happy to package it and maintain it such that a `hunspell-cy` package exists for spellchecking. I've taken a look at https://wiki.archlinux.org/title/Package_Maintainers#How_do_I_become_a_Pack… and it looks like I would need to first maintain an AUR package to achieve this, as well as find two current package maintainers willing to sponsor me? I can see a user-contributed package exists here already https://aur.archlinux.org/packages/hunspell-cy-git though it hasn't been maintained recently. I'm happy to provide an alternate package, or take over maintenance, or skip AUR completely and provide an official package: whatever is Arch's process here. I've never packaged software for Linux distros, but I am fairly active with KDE https://invent.kde.org/finw and have been providing the Hunspell data for LibreOffice https://extensions.libreoffice.org/en/extensions/show/27374 / https://github.com/fin-w/LibreOffice-Geiriadur-Cymraeg-Welsh-Dictionary, so I'm certain I can maintain the `hunspell-cy` package long-term. Can somebody help me get started? Many thanks, Finley

3 2

Name resolution failure
by Frank 14 Jan '26

14 Jan '26

I'm having troubles connecting my laptop with the internet. I can ping my router and can connect to the web interface. Connection with my phone works well (mobile data turned off) so internet connection is not an issue. Got to be something with my machine. I updated last night. I'm using NetworkManager with default setting to manage my connections. Really not sure what to look for. THX Frank

2 2

Whats going on with the mirror serving the UAE???? (mirror.hafeezh.com)
by Khalifa Alahmedi 13 Jan '26

13 Jan '26

Hey guys! ^^ One of my friends noticed that the ISO images of Arch Linux from the UAE mirror are corrupted when he tried downloading one. Checked the error logs for it on the Mirrors page of archlinux.org <http://archlinux.org/>, and there seems to be some rsync/https timing out errors. Does anyone know what could be going on? (Also ill be contacting the guy maintaining it soon) All the best, KMA! :3

1 0

Dask dashboard - Python 3.14
by Stephane Goldstein 12 Jan '26

12 Jan '26

Since last Python update 3.13 to 3.14, I get the following error when opening up the Dask dashboard on the browser: ```` Dask needs bokeh>=3.1.0 for the dashboard. Install with conda: conda install "bokeh>=3.1.0" Install with pip: pip install "bokeh>=3.1.0" ``` Dask is working fine though. Below are my installed versions: ``` $ pacman -Q python python-bokeh python-dask python-distributed python 3.14.2-2 python-bokeh 3.8.1-1 python-dask 2025.12.0-3 python-distributed 2025.12.0-2 ``` Any ideas on what could the issue be ?

3 5

Turning Bluetooth headset off crashes GNOME
by Ariadna Vigo 09 Jan '26

09 Jan '26

Hi! I've found a very curious issue and I'd like to know whether anyone else can reproduce it too. I have a BT headset which has always worked and works flawlessly. However, since last week if I turn the headset off *while connected*, GNOME crashes and bounces the system back to the GDM login screen. This does *not* happen if I disconnect the headset first and then turn the device off. I have isolated a core dump after the crash: [1]. As there have been many updates to different system components in the last couple of days, I don't have a clear idea where the problem lies in. Any possible ideas or similar experiences might be helpful to understand the issue and, if identified, submit a bug report to the corresponding project. [1]: https://paste.sr.ht/~ariadna/0361313386b759307f83d670909b1877e43ce561 -- Ariadna Vigo https://ariadnavigo.xyz gpg 0xC948873069856D6D

2 2

Hit by the pacman/makepkg "error: could not open file /var/cache/pacman/pkg/download-uq8lgk: Error reading fd 7"
by David C Rankin 09 Jan '26

09 Jan '26

All, Devs, From earlier discussions, I thought the empty stray directory problem in /var/cache/pacman/pkg had been identified and solved. But I was hit again today with directories created this year, e.g. sudo pacman -Sc error: could not open file /var/cache/pacman/pkg/download-peUZUV: Error reading fd 7 error: could not open file /var/cache/pacman/pkg/download-uq8lgk: Error reading fd 7 Where, e.g.: $ sudo ls -al /var/cache/pacman/pkg/download-peUZUV total 352 drwx------ 2 alpm alpm 4096 Jan 1 03:42 . drwxr-xr-x 4 root root 352256 Jan 8 21:56 .. I went ahead and just removed both directories. If there are any logs you are interested in, let me know, but I know the commit that led to this has already been identified. I'm just not sure of the fix. -- David C. Rankin, J.D.,P.E.

4 3