| 11 Nov 2024 |
 K900 | * People have been talking about just forking SDDM | 19:06:05 |
 πΌπ. π±ππ πππ π π·ππ | 
Download image.png | 19:11:24 |
| πΌπ. π±ππ πππ π π·ππ | has it changed? | 19:11:30 |
| πΌπ. π±ππ πππ π π·ππ | In reply to @k900:0upti.me
oomd is basically earlyoom absorbed into systemd and cleaned up really? it seemed way different to me, e.g. managing cgroups instead of individual processes | 21:28:05 |
| πΌπ. π±ππ πππ π π·ππ | which isn't a good thing in my opinion | 21:28:14 |
 mjm | why is that not a good thing? | 21:30:08 |
| πΌπ. π±ππ πππ π π·ππ | In reply to @mjm:midna.dev
why is that not a good thing? because many things don't have own cgroup | 21:48:18 |
| πΌπ. π±ππ πππ π π·ππ | and, well, I don't see what kind of advantage killing entire cgroup could provide | 21:48:36 |
| πΌπ. π±ππ πππ π π·ππ | like, if a tab in bwoser consumer too much ram, earlyoom will kill just the relevant process | 21:49:07 |
| πΌπ. π±ππ πππ π π·ππ | oomd would likely kill entire browaser | 21:49:15 |
| πΌπ. π±ππ πππ π π·ππ | * oomd would likely kill entire browser | 21:49:18 |
| mjm | a cgroup lets you say "these things are all related to the same task" and evaluate usage as an entire unit | 21:49:33 |
| πΌπ. π±ππ πππ π π·ππ | In reply to @mjm:midna.dev
a cgroup lets you say "these things are all related to the same task" and evaluate usage as an entire unit i don't need it it though when killing | 21:49:53 |
| πΌπ. π±ππ πππ π π·ππ | In reply to @mjm:midna.dev
a cgroup lets you say "these things are all related to the same task" and evaluate usage as an entire unit * i don't need it though when killing | 21:49:58 |
| πΌπ. π±ππ πππ π π·ππ | the less processes killed the better | 21:50:11 |
| mjm | idk if that's always true | 21:50:20 |
| πΌπ. π±ππ πππ π π·ππ | haven't encountered situation when it's false yet | 21:51:09 |
| πΌπ. π±ππ πππ π π·ππ | * like, if a tab in browser consumes too much ram, earlyoom will kill just the relevant process | 21:51:55 |
| πΌπ. π±ππ πππ π π·ππ | * i don't need it when killing though | 21:52:06 |
| mjm | ok | 21:52:10 |
| πΌπ. π±ππ πππ π π·ππ | maybe that's good for servers but doesn't seem to be good on desktop | 21:52:47 |
| 12 Nov 2024 |
|  @azahi:azahi.cc changed their profile picture. | 18:51:18 |
| K900 | https://github.com/NixOS/nixpkgs/pull/355517 | 21:00:15 |
| K900 | Looking to drop libinput backend from our default Qt | 21:00:27 |
| 14 Nov 2024 |
| πΌπ. π±ππ πππ π π·ππ | It's sad that nix makes leaving it enabled expensive... | 06:53:05 |
| K900 | We could technically just move the libinput plugin into its own output | 06:54:30 |
| K900 | * We could technically just move the libinput plugin into its own derivation | 06:54:45 |
| K900 | If someone really wants it | 06:54:53 |
| K900 | https://github.com/NixOS/nixpkgs/pull/355932 | 16:25:26 |
| K900 | More qtdeclarative backports | 16:25:30 |
