wicked witch of the unix

for a system that isn't meant to be manually updated without flashing a new image... they could've just used gentoo and avoided all that RE: https://floss.social/@collabora/116936104000458547

@yukijoou@fedi.kemonomimi.gay @davidgerard@circumstances.run by what i understand, yes? at least they can't sue for patent infringement i think i guess things like trademark would be another story but then you'd just have to not use "framework" in your branding i do remember seeing people making custom "modules", with open design, selling pre-builds, and even advertising those sales on the framework forums, so i don't think fw has any problem with that by itself

"i only use lower case letters" "i don't fw capitalism" so me

@fiore@brain.worm.pink @rose@pl.pinkro.se root user can read /etc/shadow, normal users can pam has a setuid helpers to do that inbehalf of the user

@fiore@brain.worm.pink "upstream control" "distros bad" "users can't be trusted" etc etc not their actual arguments but basically what those arguments mean

@snacks@netzsphaere.xyz @rose@pl.pinkro.se this was milk coffee, so the mouthfeel was already not just-coffee -- i drink it like that cuz i can't stand any bitter flavor idk how "dialed in" my coffee is but i just do ~1:14 ratio with some preground coffee i like

@fiore@brain.worm.pink it's a common sentiment around gnomeware, specially after the bottles drama and it's one of the two reasons i stay away from the ecosystem

imho "users should check the diff!" is kind of insane cope to justify not having a slither of moderation meanwhile gentoo's guru manages to provide a user repository where committed ebuilds are at least reviewed over-the-top before merging it's still not fully trustworthy, so it does get ~keywords (meaning each package from guru needs to be explicitly accepted), which is good, as "users shouldn't ultimately trust" doesn't mean "there's no need for any kind of moderation"

@emmy@transfem.social a patch for it was applied in in 05-04-2024 so it should be in 0.5.13 -- thing is we had 0.5.12 since 2022, it went 4 years without a release... so a lot of distros don't have 5.13 yet

@SRAZKVT@tech.lgbt @dillo@fosstodon.org @DLC@mstdn.social

the nav tag doesn't really have much special semantics though, it's mostly just there to say "hey this is navigation links", so css would still be needed

and that's the bit that should be fixed, instead of patching it with css

in this case, give nav enough semantics that the user's browser style can render it in a way that makes sense (menu on phones, bar on desktops/tablets, etc)

battle against a true hero

@coolbean@brain.worm.pink @fiore@brain.worm.pink actually no this is a bad idea it makes cross compilation hell hm, what does meson and autotools do, lemme look it up

smartcard reader framework 13 "expansion card"

i wish the cli of the rc-* tools in openrc had been thought out better from the start as it stands they're both not great, and hard to change without it being a breaking change

@dalias@hachyderm.io @hailey@hails.org i right i forgot about that then yeah disabling netlink will have a major impact on most DE tools that interact with udev in some way

@djchateau@infosec.exchange enough that i've by now seen multiple people migrating away from it due to a big influx of bugs and regressions in newer releases (which didn't use to be common before the slop) (this is about bitwarden, keepassxc idk)

@hailey@hails.org @dalias@hachyderm.io unless you're already running xorg/a wayland compositor as a unit, that would still be an issue though this LPE doesn't really matter for personal systems so the only point i'd worry about is systemd-logind since that is still active even on server installs

This is by design and won't be changed. If you want to authenticate other users by a non-root PAM process you have to change your /etc/shadow permissions to allow reading it by the non-root PAM process. For example have the non-root PAM process run with some special group and make the /etc/shadow readable by this group. The unix-chkpwd is setuid to only allow reauthentication of the user that is calling it such as in case of the screensaver screen lock.

...

it's by design

okay I GUESS

RE: https://social.vlhl.dev/objects/dbf03a43-db43-45b6-834f-8d26899dd603

@dalias@hachyderm.io @hailey@hails.org killing AF_NETLINK would also kill uevents for udev which matters less on servers i suppose but i have no idea how systemd systems would react to that

oohh there's a store in frankfurt that sells latin american good coxinha, kibe, paçoca, bis, guaraná, yoki flour, gah i miss those have to go there sometime...

@dalias@hachyderm.io @whitequark@social.treehouse.systems @mcc@mastodon.social not just glibc, there quite a list of libraries they explicitly "assume to be present on the host system" https://github.com/AppImageCommunity/pkg2appimage/blob/master/excludelist

We may add DNS SRV records as supported mechanism in the future, but we currently do not.

thunderbird

fuck you

@rose even when there's info on the page (gnu sed), the posix manpage is usually a lot more detailed too though

since i don't code in my server, i could replace nvim with a simpler vi impl i think, which would have less dependencies, and as a bonus ain't slopped

kms

so, bitwarden is badly slopped now, keepassxc too, 1password gives money to fascists (omarchy), and by a quick glance lastpass is also full on AI (if you could even comsider trusting it, has plenty of data leaks before) are unix pass(2) and keepasschi (a keepassxc fork), the only surviving safe password manager?

https://github.com/OpenRC/openrc/releases/tag/0.64 openrc 0.64 out, highlights being RC_PATH and reexport dependencies

@fiore@brain.worm.pink @coolbean@brain.worm.pink (oh i was already typing a reply) first, argc itself purely is defined by the abi, e.g. for the x86_64 sysv abi¹ (which linux uses), the argument count is defined as "an eightbyte", which is more than an int in x86_64 (4 bytes) but c does require main to be `int argc, char *argv[]`, so, c spec time, does int have a size? yes... kinda², the standard says that UINT_WIDTH, the width for an object of type unsigned int, is 16 bits or more so your minimum, is 2 bytes, though in most platforms, it's 4 bytes, and "natively" outside c in x86_64 sysv specifically, it's actually 8 bytes, because it's just the word size of the processor -- your libc then implements _start, parses the aux data in the stack, and calls your main function passing in the value as an `int` not that it matters much, tbh, even with the lowest range here, 2 byte int, that's still 32767 arguments, you're gonna hit posix's ARG_MAX wayy before you hit that limit 1: https://refspecs.linuxbase.org/elf/x86_64-abi-0.99.pdf -- figure 3.9, initial process stack 2: https://open-std.org/JTC1/SC22/WG14/www/docs/n3220.pdf -- 5.2.5.3.2 - Characteristics of integer types

@dalias@hachyderm.io @neal@social.gompa.me @datenwolf@chaos.social @miah@hachyderm.io fwiw the current wip build of my compositor does software rendering only (wl_shm buffer composited on the cpu), and all gtk programs i've tested worked fine so at least so far it's still a thing to not need a gpu, unless gtk is rendering on the gpu and copying the buffer to main memory which would be uhhh why?? but also not a problem with wayland

@coolbean@brain.worm.pink @fiore@brain.worm.pink either keep a list of sizes per platform, or better imo, during build time build a simple c program that just returns sizeof(int)

@fiore@brain.worm.pink yes yes

drip coffee machines still require the user to measure out the coffee, put in a filter, clean up honestly what's the point of the machine by then it's just a fancy water boiler with a coffee pot bolted on /j

@coolbean@brain.worm.pink @fiore@brain.worm.pink well size of argc is int, size of int depends on the system abi the problem here is getting the size of int from not-a-c-compiler tbh though, how does your interop works, is it calling forth from c, or calling c from forth, or smth else?

@dalias@hachyderm.io @hailey@hails.org iirc udev rebroadcasts a (modified) event in another netlink layer for libudev (user) programs to consume

my next server will be a gentoo musl install :3

normal distros: mirrors distributed around the globe, usually hosted by universities or other contributors donating server storage and network, resulting in resilient package distribution

ubuntu:

RE: @joeress@mastodon.social

gentoo forks in the wild

RE: @aboutzoos@mastodon.social

suckless already removed the entry of "we distances ourselves from fascism" from the faq that really didn't last long, i can't even

@SRAZKVT@tech.lgbt @DLC@mstdn.social @dillo@fosstodon.org i feel like the majority of things that need css for layout should be structually semantic instead e.g. it doesn't make sense that a navigation side/top bar is a div holding a unordered list, that can only show that it's actually supposed to be if the css is correct yea exists iirc, but i see people not using it often -- imo if we don't allow much css at all, it'll mean people can't hack around having proper structure

@coolbean@brain.worm.pink @fiore@brain.worm.pink if the function takes argc and argv, then you can just define the argc of that function to a known sized type, like int32_t or int64_t or whatever word size forth has, and let the compiler do the conversion or, if it's that dynamic, yeah, take sizeof(int) as an argument somewhere

i understand, mechanical hands, are the ruler of everything

honestly bless the posix manpages

man yacc > "please look at info" man yacc.1p > all info right there

@davidgerard@circumstances.run and the patent seems to even have gotten rejected (then abandoned) which, lol

rose struggled to post this weird ass blank keyboard