Posts
2239
Following
155
Followers
75
AS4242423219 on DN42
Also @noisytoot@mice.tel in case chinchillas eat the cables
I should publish my regreet patches somewhere (or send them upstream, but the patch adding a suspend button definitely isn't ready for that since it doesn't make it optional)
0
0
0
Edited 3 months ago
sddm is broken in alpine (again) so I just switched back to greetd+regreet and fixed the weird large cursor issue (which involved adding support for setting the GTK cursor-theme-size option) and figured out how to hide the titlebar (passing -d to cage to disable CSD)
1
0
0
re: genai, github
Show content
@sugar @lumi @kemona_halftau I thought the reason they rewrote bun to rust was that zig refused to accept their slop PRs?

anyway, this code is completely unreadable: https://github.com/vercel-labs/zero/blob/dbe552787bd1079f740296aed3d7f3b53f3f4e81/compiler-zero/src/wasm.0

what makes it "agent-first" seems to be that no human would ever want to touch this
1
0
4
@kemona_halftau report the bug and temporarily remote-suspend the user and delete the posts to prevent it from crashing until the bug is fixed?
0
0
1
@domi @swirly I think it's just for backtraces and they didn't know that musl doesn't have it so the build system doesn't check for it and just assumes that it exists
1
0
0
@domi @swirly the package isn't updated to supertux 0.7 yet, which doesn't build on musl because it depends on the glibc-specific execinfo.h. there's an implementation of that that works on musl (libexecinfo), but it's not packaged yet either
1
0
1
re: rsi
Show content

@kemona_halftau I currently just use emacs with mostly-default keybindings, but maybe I should change that and learn evil-mode. I find modal editing weird but maybe that’s just because I’m not used to it. The defaults aren’t really designed for modern keyboards, they’re designed for a space-cadet keyboard, which has a load of extra modifiers (which emacs doesn’t use by default but does support) and has the positions of ctrl and meta swapped from modern keyboards (meta = alt on modern keyboards, and what emacs calls alt doesn’t exist) which is why some of the default keybindings seem awkward.

I also want to get some more digital piano pedals and use them as extra modifier keys with a pi pico to translate to USB, but I couldn’t figure out how to get TinyUSB to simulate a HID keyboard (it might be easier to just use USB serial and have a userspace program that sends the keys with uinput)

0
0
0
re: rsi
Show content
@kemona_halftau it is by default, but it doesn't have to be. there's evil-mode for example if you want vi keybindings
1
0
1
@whitequark @dalias @wwahammy LPEs are certainly an issue (although they're also an issue for any CI that doesn't use proper VMs), but Nix doesn't just allow any random unprivileged user to configure a substituter, right?
1
0
0
@whitequark @dalias @wwahammy if a malicious Nix builder can poison the cache, wouldn't that mean that a multiuser Nix system is insecure as well, since unprivileged users are allowed to build and install packages?
1
0
1
@kemona_halftau but emoji reacts are cool. and some day maybe I will implement something cursed like emoji reacting to emoji reacts so conversations can be done entirely in emoji reacts
0
0
0
@kemona_halftau I'd miss emoji reactions on gotosocial (it doesn't support them yet)
1
0
0
@hexaheximal @justsoup in order to get sensible indentation behaviour with tabs in emacs, you need smart-tabs-mode and for tab-width to be equal to the c-basic-offset (or equivalent). the problem with the latter is it means you can't enforce sane indentation behaviour via mode-line comments (the "-*- mode: c; -*-" thing) or .dir-locals.el without either also enforcing a default display tab width or making it sufficiently complex that emacs will ask you whether you really want to load the .dir-locals.el because it might execute arbitrary code (you can't reference other variables without that)
0
0
0
Show older