06:48 <leviathan> ifreund: Yes, you're right. That's exactly how I understood it initially too.

06:48 <leviathan> But in the process of my implementation, I found that no matter what value I send using text_input_rectangle, it doesn't affect the processing of fcitx5 at all.

06:53 <leviathan> I only need to create the scene tree and set the position of the node, and the fcitx5 popup surface appears in the expected position. I don't understand what role text_input_rectangle is playing.

07:28 <LarstiQ> adamcstephens: finally managed to update to 0.3.0, had to overlay zig_0_11 with a patch. Soo happy to have `send-to-output -current-tags` again \o/

10:39 <lordmzte> How do I completely disable GTK4 client-side titlebars? I already have some custom CSS as a workaround (https://git.mzte.de/LordMZTE/dotfiles/src/branch/master/.config/gtk-4.0/gtk.css.cgt) and this does work for most applications, but as of a recent update, it doesn't appear to be working for ghostty (the close button still shows). Is there

10:39 <lordmzte> something I'm missing in this stylesheet, or is this something ghostty needs to fix?

11:12 <ifreund> lordmzte: Ghostty could almost certainly add an option to disable CSD. I don't know anything about gtk4 workarounds unfortunately, there is some stuff on the river wiki though

11:12 <ifreund> in other news, Zig 0.12.0 is out!

11:13 <ifreund> And I just tested building river with the self-hosted x86 backend and it just works :)

11:13 <novakane> oh, unnecessary use of var is pretty cool

11:14 <ifreund> it's fast enough that it feels instant compared to the "LLVM Emit Object..." step

11:14 <ifreund> so that's very cool

11:15 <novakane> ifreund: are you gonna start tagged release deps with zig 0.12 to use with the package manager?

11:15 <novakane> how do I tried the self hosted backend, damn I really followed nothing in this release

11:16 <ifreund> novakane: it tells you how here: https://ziglang.org/download/0.12.0/release-notes.html#x86-Backend

11:17 <ifreund> I don't know if I want to switch to the package manager yet or not...

11:17 <novakane> thanks, didn't make it to this part of the change log yet

11:18 <ifreund> the distro packaging story is OK with the package manager but not perfect

11:18 <ifreund> https://github.com/chimera-linux/cports/pull/1395

11:21 <ifreund> In any case, I won't be merging 0.12.0 support to master just yet. I want to give some more time for bug fixes for 0.3.1 to accumulate before I make backporting harder on myself

11:21 <lordmzte> ifreund: I did take that CSS from the river wiki and also set GSettings accordingly, I guess it must be an issue with ghostty then. thank you!

11:22 <ifreund> (0.3.1 will be Zig 0.11, 0.3.2 will probably happen shortly afterwards and just upgrade the zig version)

11:29 <novakane> fair, not like we couldn't wait for 0.13 or something to use the package manager

12:07 <adamcstephens> LarstiQ: what was the zig_0_11 patch you needed?

13:35 <LarstiQ> adamcstephens: this unupstreamable bit: https://github.com/par5er/zig/commit/b5a82720d308329c9766b48ce85d371a37995a5c

13:36 <LarstiQ> ifreund: ooh, time to read the 0.12 release notes

13:51 <waleee> I don't think I have seen a wasm-based library documentation outside of https://www.uiua.org/ before

13:52 <waleee> (zig's new autodoc)

16:01 <adamcstephens> LarstiQ: ahh yes the Apple silicon page size

16:05 <adamcstephens> when will that be resolved by zig?

16:11 <ifreund> when someone implement runtime page size detection and updates the standard library correctly

16:18 <adamcstephens> 👍

19:31 <LarstiQ> now that 0.12 is out perhaps someone has a spare moment for it, but I can manage like this. Just had some Nix learning to do before I wrapped my head around how one is supposed to do this