[2026-01-06T01:01:08Z] fultilt: https://www.nytimes.com/2026/01/05/business/media/corporation-for-public-broadcasting.html [2026-01-06T01:57:53Z] is there any body using librewolf? how did you compile it? [2026-01-06T01:59:51Z] me elisoli [2026-01-06T01:59:53Z] using flatpak :P [2026-01-06T02:02:07Z] I don't see why it should be much different from building firefox tho [2026-01-06T02:23:11Z] uh, I think I found a bug in the mdev-like-a-boss config we're using [2026-01-06T02:23:37Z] `cpu([0-9]+)` matches to nothing on my machine, it's basically ignored [2026-01-06T02:24:12Z] busybox mdev (at least) straight up detects and makes `cpu/$ID/cpuid` [2026-01-06T02:24:39Z] I wonder if the kernel changed or something [2026-01-06T02:35:17Z] i am not an expert tbh, although i know a lot about programming (im a dev xd) tho i can't understand nothing in that firefox build script [2026-01-06T02:35:58Z] librewolf when compilng have me weird errors with python packages failing inside rust compilation processes O_o [2026-01-06T03:00:05Z] perhaps they added some other dependency [2026-01-06T03:02:49Z] got my noctua coolers in [2026-01-06T03:03:08Z] libreboot claimed am3 coolers would work with the socket c32 mount [2026-01-06T03:03:17Z] noctua cooler supports am3 officially [2026-01-06T03:03:23Z] noctua cooler does not mount to the c32 bracket [2026-01-06T03:03:39Z] noctua *used* to make a secufirm2 kit to allow mounting to c32 [2026-01-06T03:03:49Z] but they no longer make it and i can't find any in the wild [2026-01-06T03:03:50Z] shoot me [2026-01-06T03:39:03Z] midfavila: are you using libreboot [2026-01-06T03:42:09Z] going to be, yes [2026-01-06T03:42:29Z] at some point i really need to look at how their build system works so i can do it myself [2026-01-06T03:42:36Z] not exactly a hot take but i'm not a leah fanatic [2026-01-06T03:44:42Z] it's whatever [2026-01-06T03:44:48Z] gonna be returning these coolers i think [2026-01-06T03:44:59Z] they don't seem compatible in any way shape or form with socket c32 [2026-01-06T03:45:02Z] very disappointed [2026-01-06T03:53:47Z] aight news flash [2026-01-06T03:53:55Z] busybox mdev is actually very dumb [2026-01-06T03:54:37Z] The Isopod [2026-01-06T03:54:39Z] I think it scans devices straight up wrong, according to the kernel docs [2026-01-06T03:54:59Z] like, when it does its first scan, a "coldplug" it straight up misses a LOT of devices [2026-01-06T03:55:14Z] and it's not like it does not like them, it reports their events fine [2026-01-06T03:55:17Z] case in point, backlight stuff [2026-01-06T03:55:31Z] mdevd works correctly [2026-01-06T03:56:13Z] so uh, I might *need* to switch to it if I want something "correct". Either that, or I need to resort to hacks like I always did to achieve what I want (user writable backlight devices and such) [2026-01-06T03:56:29Z] also, for reference, this is what it should do: https://www.kernel.org/doc/html/latest/admin-guide/sysfs-rules.html [2026-01-06T03:56:49Z] to be clear, I haven't checked the code and don't really plan to, so it's really just a guess at why it misses so much stuff [2026-01-06T03:57:40Z] also TIL that they're planning to add a new /sys/subsystem directory with every single device in there [2026-01-06T03:58:28Z] oh wait everything is (going to be?) under /sys/devices, with a new /sys/subsystem directory for filtering by classification [2026-01-06T04:12:50Z] alright coolers are all boxed up [2026-01-06T04:12:59Z] c32 is a socket i'm unfamiliar with [2026-01-06T04:13:00Z] damn i was super excited for them too [2026-01-06T04:13:05Z] mid what cpu is in this thing [2026-01-06T04:13:18Z] opteron 43xx something or other [2026-01-06T04:13:20Z] it's a piledriver [2026-01-06T04:13:24Z] o [2026-01-06T04:13:27Z] or, well, they're piledrivers [2026-01-06T04:13:30Z] i forgot opterons existed [2026-01-06T04:13:35Z] those are so cheap now [2026-01-06T04:13:37Z] C32 is the workstation socket from the era [2026-01-06T04:13:47Z] G38 iirc is the server equivalent [2026-01-06T04:13:57Z] im surprised you didnt go with like a xeon x5650 setup or something [2026-01-06T04:13:58Z] both obsolete the previous F-series [2026-01-06T04:14:09Z] this is the main desktop board for libreboot [2026-01-06T04:14:23Z] it's also relatively affordable [2026-01-06T04:17:08Z] im curious to see how good you find it [2026-01-06T04:17:16Z] iirc opterons aren't exactly fast [2026-01-06T04:17:20Z] and weren't at the time [2026-01-06T04:17:31Z] 3.8ghz [2026-01-06T04:17:37Z] twelve cores [2026-01-06T04:17:38Z] not bad [2026-01-06T04:17:41Z] not outstanding [2026-01-06T04:17:52Z] well i figure you arent gaming on it so i figure itll be fine [2026-01-06T04:18:01Z] yeah it's mostly for multicore stuff [2026-01-06T04:18:24Z] i need a set of calipers [2026-01-06T04:18:29Z] there's so much random shit i really need to get [2026-01-06T04:18:34Z] calipers, exacto knives, etc [2026-01-06T04:18:34Z] me [2026-01-06T04:18:38Z] i have an entire list [2026-01-06T04:18:50Z] i need calipers to measure bits of this sk8855 board so i can make a case for it [2026-01-06T04:19:28Z] kris_: https://www.kickstarter.com/projects/omni-r/omni-r-multifunctional-portable-ruler [2026-01-06T04:19:32Z] bam [2026-01-06T04:19:36Z] there's your everything [2026-01-06T04:19:44Z] i make no claims as to its quality but it sure is nifty [2026-01-06T04:22:24Z] i mean a set of calipers is like $8 [2026-01-06T04:22:30Z] who knows how accurate they are but yeah [2026-01-06T04:24:50Z] probably more than enough [2026-01-06T04:25:14Z] bigclive did a teardown of those extremely cheap digital calipers and IIRC he was surprised of the precision [2026-01-06T04:25:30Z] really clever lil gadgets [2026-01-06T04:27:18Z] i need to make a fortune expansion thingy that contains all of the quotes from dracula flow [2026-01-06T07:51:51Z] dery: who even needs mdev [2026-01-06T07:52:42Z] i feel like playing a game but idk what i want to play [2026-01-06T07:52:47Z] true woes [2026-01-06T08:02:04Z] sad_plan: I do [2026-01-06T08:02:24Z] like, joysticks and stuff [2026-01-06T08:02:37Z] i must consoom and things like that [2026-01-06T08:02:43Z] btw [2026-01-06T08:02:50Z] https://dwheeler.com/essays/filenames-in-shell.html nifty resource, never heard of it before [2026-01-06T08:03:16Z] dery: you're more pure than me [2026-01-06T08:03:18Z] i use udev [2026-01-06T08:06:15Z] that's perfectly fair tbh [2026-01-06T08:06:22Z] like, the apis are dirty down there fr [2026-01-06T08:06:34Z] really i use it because i'm too lazy to not [2026-01-06T08:06:45Z] I find it really satisfying to bang my head at this stuff tho [2026-01-06T08:06:49Z] its kinda baked into one of voids core services, i could fix this pretty easily but im lazy [2026-01-06T08:07:07Z] yea it's really no biggie [2026-01-06T08:07:29Z] especially as then certain software won't work without libudev-zero [2026-01-06T08:07:43Z] and even then it's got some stubs so stuff *will* break [2026-01-06T08:14:58Z] miau [2026-01-06T08:16:07Z] indubitably [2026-01-06T09:11:33Z] dery: I dont belive you actually need mdev for that. you just need to fix the permissions yourself C: [2026-01-06T15:05:34Z] sad_plan: I did [2026-01-06T15:06:02Z] but I was looking into why I could not manipulate backlight stuff from mdev, which I should, I think [2026-01-06T15:06:12Z] you can from udev [2026-01-06T15:12:56Z] dery: I see. [2026-01-06T15:34:28Z] also apparently mdev-like-a-boss is built for... mdevd? there's a bunch of `cpu` regexes which I can't get to trigger on busybox mdev but do on mdevd [2026-01-06T15:35:01Z] I'm genuinely starting to think that perhaps a kernel update changed the /sys/ directory structure and something broke along the way [2026-01-06T15:47:28Z] perhaps. cant you just script the correct permissions with a cronjob instead? [2026-01-06T15:52:32Z] again, I did in a `.boot` script previously [2026-01-06T15:52:45Z] that works 100% [2026-01-06T15:53:04Z] I was simply playing around and found those (undocumented?) limitations of busybox mdev [2026-01-06T15:53:44Z] right. i mustve missed that [2026-01-06T15:53:50Z] I think it looks ever so slightly cleaner (and correct) to do it on device addition instead of bootup but on my laptop it's 100% irrelevant as both the drm adapter and the cpus aren't going anywhere :P [2026-01-06T15:54:09Z] no worries sad_plan [2026-01-06T15:55:54Z] im in favor of choosing w/e solution works, while also being the simplest one. so use w/e works basically :p [2026-01-06T21:09:30Z] phoebos: any idea how to contact hovercats ? I need an update for nettle in community for gnutls [2026-01-06T21:09:45Z] but I expect he isnt' active anymore [2026-01-06T21:50:56Z] the package might need to be orphaned if they *are* inactive. maybe for now just bump the version on a local fork? [2026-01-06T22:10:24Z] what do you mean by "local fork" ? [2026-01-06T22:24:21Z] like make a copy of the package you need updated, make the changes yourself *to* that copy, and then you can use that until upstream is updated [2026-01-06T22:24:57Z] usually all you need to do to update a package is point the source file at the new upstream source, bump the version file to reflect that new source, and regenerat the built package's checksums [2026-01-06T22:27:53Z] regenerate, even [2026-01-06T22:30:24Z] ok, I get it. I want to update gnutls on community, because I'm the maintainer of gnutls. I can wait. But at some time it needs to be updated. [2026-01-06T22:32:03Z] and to orphan nettle, sincerly I have no idea how to contact the maintainer. He has an email ending with "@noreply.codeberg.org". I assume it's not usable. [2026-01-06T22:32:19Z] I could open an issue on community. Maybe it's the best. [2026-01-06T22:39:14Z] yeah [2026-01-06T22:46:01Z] thank you! [2026-01-06T22:47:24Z] np [2026-01-06T22:48:35Z] ngl thinking i'm gonna quit the bike job... kinda got what i wanted out of it. vancouver is looking really nice right now too