2026/03/12

Newest at the top

2026-03-12 07:28:41 +0100vados(~vados@46-133-140-234.mobile.vf-ua.net)
2026-03-12 06:58:05 +0100vados(~vados@46-133-57-88.mobile.vf-ua.net) (Ping timeout: 252 seconds)
2026-03-12 03:58:20 +0100 <geekosaur> (if I got the math right)
2026-03-12 03:58:11 +0100 <geekosaur> (fwiw the repeat delay is in ms. if it somehoww got set to MAX_INT, that's around 23 days)
2026-03-12 03:43:28 +0100 <geekosaur> it's also possible that the repeat delay is set really high (like, minutes or even hours) or that the repeat mask has been cleared somehow (all 0s in "auto repeating keys", which is likely to be rather difficult to fix)
2026-03-12 03:33:59 +0100 <geekosaur> (or "xset q" to see what the settings including key repeat settings are)
2026-03-12 03:33:34 +0100 <geekosaur> try: xset r on
2026-03-12 03:15:08 +0100 <haskellbridge> Is there something wrong with my laptop and/or my Linux settings?
2026-03-12 03:15:05 +0100 <haskellbridge> <i​qubic (she/her)> So, for some reason, my laptop's built-in keyboard has stopped doing key-repeat. No matter how much I hold down a key, Linux only ever registers it a single time.
2026-03-12 00:21:47 +0100stackdroid18(~stackdroi@user/stackdroid) ()
2026-03-11 21:45:54 +0100scardinal(~supreme@0x573d64a9.static.cust.fastspeed.dk) (Quit: leaving)
2026-03-11 21:42:14 +0100 <geekosaur> resizing it manually afterward eorked fine, but next time you started your X session it would do it again
2026-03-11 21:41:48 +0100 <geekosaur> this had the amusing result that, back in the days when it had a frame size bug, it would get into a fight with kdewm that led to its initial window size visibly stepping down to about 3x2
2026-03-11 21:40:06 +0100 <geekosaur> emacs has a certain tendency to assume that the window manager (or for that matter anything else) should be subordinate to it
2026-03-11 21:39:06 +0100 <geekosaur> or reconfigure it with emacs commands instead of the mouse
2026-03-11 21:38:42 +0100 <geekosaur> I don't have an emacs confoiguration that saves frame information; if you do, you might need to turn it off or manually edit the save
2026-03-11 21:25:20 +0100 <geekosaur> those together (especially the second) sound like it's emacs doing it and the most you can accomplish from xmonad is starting a war with it (that is, you get a resize loop)
2026-03-11 20:31:06 +0100stackdroid18(~stackdroi@user/stackdroid) stackdroid
2026-03-11 20:04:19 +0100ectospasm(~ectospasm@user/ectospasm) ectospasm
2026-03-11 19:54:51 +0100Enrico63(~Enrico63@host-82-61-84-117.retail.telecomitalia.it) (Quit: Client closed)
2026-03-11 19:37:37 +0100ectospasm(~ectospasm@user/ectospasm) (Quit: WeeChat 4.8.1)
2026-03-11 18:53:13 +0100 <haskellbridge> <e​ldritchcookie> it also doesn't allow me to manually fix the window with the mouse.
2026-03-11 18:52:12 +0100 <haskellbridge> <e​ldritchcookie> i thought it was something i did wrong but i tested with another application and it works fine, when i use doFullFloat on emacs it covers more than the whole window vertically on the bottom and misses a little horizontally on the right. Are there workarounds for correctly floating emacs?
2026-03-11 17:52:32 +0100 <liskin> Even if, it'd look different than what X does
2026-03-11 17:52:17 +0100 <liskin> Yeah but maybe use the brain a bit? We're not trying to make a generic WM interface...
2026-03-11 17:50:26 +0100 <geekosaur> there was even a proposal to directly support it that got shot down for "security reasons"
2026-03-11 17:49:57 +0100 <geekosaur> the reason people miss it is wayland docs and devs tell you that's insecure and you shouldn't do it
2026-03-11 17:37:08 +0100 <liskin> I mean actually maybe that's good. Make me angrier and I might just sit down and fucking do this at last.
2026-03-11 17:36:24 +0100 <liskin> Sorry for the caps but how does everyone keep missing this bit? Makes me furious.
2026-03-11 17:36:04 +0100 <liskin> You REALLY FUCKING DON'T WANT TO HAVE TO RESTART YOUR WHOLE COMPOSITOR
2026-03-11 17:35:29 +0100 <liskin> Xmonad is configured by compiling - reload of configuration means restart.
2026-03-11 17:34:59 +0100 <liskin> (the most I managed was just under 2 years - that's completely unrealistic in the Wayland world)
2026-03-11 17:34:30 +0100 <liskin> In X11 you can restart xmonad a million times while still letting the X server run for months or even years
2026-03-11 17:34:00 +0100 <liskin> Also there's this big fundamental limitation of Wayland design - if the compositor crashes or needs to restart, your whole session goes down
2026-03-11 17:33:05 +0100 <liskin> Even Rust folks struggled to get wlroots bindings to feel good, that's why smithay is thriving these days. It's just better to do a Rust native Wayland library
2026-03-11 17:32:23 +0100 <liskin> And also it's much harder to write/maintain bindings to C APIs
2026-03-11 17:31:59 +0100 <liskin> Haskell is garbage collected - so that means latency might suffer unpredictably
2026-03-11 17:31:01 +0100 <liskin> 2 years ago that would mean being limited and being behind the protocols supported by gnome/kde/wlroots, but I think these days smithay is fine
2026-03-11 17:30:35 +0100 <haskellbridge> <E​nrico Maria De Angelis> What's fundamentally flawed with doing it in Haskell (out of curiosity, and with zero prerequisites to understand the answer, probably).
2026-03-11 17:30:21 +0100 <liskin> Smithay (the Rust alternative to wlroots) being in a good shape these days means doing it in Rust should be fine
2026-03-11 17:29:41 +0100 <liskin> I've been saying for years - let's just do that in Rust or C, and let it rpc to a Haskell window manager.
2026-03-11 17:29:05 +0100 <liskin> I still don't know why anyone's trying to have the compositor itself be written in Haskell.
2026-03-11 17:15:26 +0100vados(~vados@46-133-57-88.mobile.vf-ua.net)
2026-03-11 17:01:59 +0100Enrico63(~Enrico63@host-82-61-84-117.retail.telecomitalia.it) Enrico63
2026-03-11 16:02:44 +0100deepy(deepy@user/deepy) deepy
2026-03-11 16:02:17 +0100deepy(deepy@user/deepy) (Read error: Connection reset by peer)
2026-03-11 13:29:32 +0100DigitteknohippieDigit
2026-03-11 12:59:25 +0100Digitteknohippie(~user@user/digit) Digit
2026-03-11 12:35:48 +0100Digitteknohippie(~user@user/digit) (Ping timeout: 246 seconds)
2026-03-11 12:24:00 +0100ChubaDuba(~ChubaDuba@46.147.102.235) (Ping timeout: 255 seconds)