>>120
I agree.

>>119
well i mean a lot of systemd is worked on by redhat but i mean even just a regular user of linux it's so much better than the other "alternatives" that there's not much competition in my mind and the "all sorts of unrelated functions in itself" is part of why it's so useful

>>122
it ain't the 70s anymore and the Unix philosophy isn't that important especially outside of low-level areas (i.e kernel space)

>>123
i think this idea comes from the fact that linux generally is terrible at "unix philosophy" in general, not because "unix philosophy" isn't good or useful. like /dev/ and /proc/ aren't really useful, except through specific command interfaces, you can't just echo and read from them to do useful things for the most part. shell scripting is fractured, every service invents it's own config methods some even inventing their own bespoke language for doing so. some distros try to be cleaner about it but there's only so much that can be done without ripping out the whole foundation and starting from scratch. Even the original unix team knew that the unix kernal was not able to maintain it's "unix philosophy" in a distributed network environment, which is why they threw it away and started plan9.

plan9/9front is like the only thing that actually does unix philosophy and the ease it provides when actually embraced from top to bottom is incredible. All the problems plan9 could be solved by forcing the unix philosophy on more hardware, not by aquiessing to the non-unix demands of the hardware. for example, in a fully plan9 world, i could from my laptop pipe the display of my tv connected to a blue ray player to my phone. instant reverse screen cast with nothing but file pipeing. of course we don't live in the world where it can run on all hardware, and where device drivers are borderline impossible to create without vendor support.

>>124
(not who you were replying to)
people often say this but, not to say plan9 is bad or anything, there are plenty of formulations of linux which can fit quite nicely into the unix philosophy way of doing things. perhaps it's lacking in the networking department sure, I'll grant that, but that doesn't mean it should be dismissed entirely. KISS linux is the obvious example of a unix philosophy adherent distro, although it's not very commonly used, the philosophy of extremely minimal system with lots of small shell scripts tying everything together is very unixy in a good way. You can see the benefits of this system because it was designed to be maintainable by a single user, and when the lead maintainer left to become an olive farmer in greece or whatever without telling anyone, everything was maintained just fine. You can set up pretty much any linux (or BSD for that matter) to run with this kind of setup, very minimal with lots of small shell scripts tying things together, and it works very well in my experience.

>>123
the unix philosophy is not now nor was it ever particular relevant to kernel space. it was primarily intended for userland programs. people simplify the unix philosophy to "write programs that do one thing and do it well" but they're forgetting the two other pieces: "Write programs to work together. Write programs to handle text streams, because that is a universal interface."

>>126
this isn't strictly correct. the kernel maps devices into userspace. the "unix philosophy" then demands that it should do so in a way that works with streamed text interfaces, ie "everything is a file"

>>127
plan9 philisophy is superior to that, why lunix lags behind & so backwards? are it's developer-architects retarded?
Let's ask Linus (from Tech Tips)!

>>128

are it's developer-architects retarded?

Yes

>>130
what the actual fuck is this shit

>>131
no fucking clue, it'd take someone way braver or stupider than me to click that link to maybe find out (and or be added to some kind of watchlist)

>>130
>>131
>>132

@mail.com

I'm an email provider! ahh email provider

>>133
What was it? You guys saved it so I could see it, right? You're not afraid of links in CURRENT_YEAR + 10, right? Right??

>>134
it was literally just an ip address

My laptop is now OS-less and I have no other computer to make a bootable USB with. Oops!

>>136
How are you postig this

>>137
I don't know.

>>137
A phone, most likely

>>137
I was on my Apple iPhone by Apple. My laptop is no longer OS-less, though! Back on Microsoft Windows 11 by Microsoft!

>>140
what a classic blunder. if the computer is already osless there is nothing left to lose, so there's no better time to move to templeOS

based templeOS user spotted

>>142
I hate how templeOS and Terry Davis are just a butt of jokes. please give some respect man

>>143
wdym...i genuinely love terry davis and i want him to come back :(

systemd haters are feeling vindicated right now

>>145
who would have possibly imagined the corporate stoogie literal bad actor trying to add everything into the monolithic virus "init system" would immediately kowtow to retarded laws and add even more bullshit into the monolith?

>>145
well they still control both x and wayland so what now, i guess we're moving to xlibre? or that guy who's rewriting x in zig? or are we just gonna be terminal multiplexing on tty1?

>>147
we should've never left the tty.

>>145
what are you referring to?
>>147
what also xlibre is ass

if it's the "Age verification" thing it's because they're being legally forced to? because multiple governments keep passing laws (because companies keep lobbying them) forcing them to do this shit? none of that info in systemd is sent anywhere anyway and it's still an optional field. maybe don't just look at news headlines and foss grifters opinions and actually read into it?

we're just having fun hating on IBM while posting from an IBM thinkpad ok !

I love following retarded laws!

Why is the path in my 64 bit operating system windows 11 capped to a character limit of 32767 (16bit signed int limit)
>>152
shut up man

>>115
>>122
Systemd is an evolution upon sysv/shellscripts and better than init systems like runit or openrc but it's a software suite with too much shit nobody uses. All the good features have already been implemented by modern init systems like s6, dinit (even has systemd like syntax) or GNU Shepherd on Guix.
Only problem is that those are less universal. You can get all on Artix (and Arch is one of the most popular distributions though most software is under AUR which isn't ideal) expect Shepherd which is pretty much specific to Guix. Chimera Linux is probably my favorite of the non-functional distros if you don't need glibc (Alpine Linux's Wiki has some resources on running glibc software using chroot).
>>153
It's only an ameriKKKan and Brazilian law. There's no reason to add it to an init system used internationally. Just do a baby_mode ISO for USA and Brazil. Honestly they're just playing into the schizophrenic's delusions before it just connected to Google's DNS I think but so does LineageOS and alot of software. I don't remember why but I guess systemd is also a DHCP client.

>>147
>>149
Xlibre is pretty much just Xorg but updated. So far they added namespaces which should add security but I'm not sure how effective it is along with some other features.
For me I'd just want TearFree with modesetting for TearFree Xorg without a heavy compositor or falling back to the old intel driver. Xorg beats Wayland in performance but not with these heavy compositors like Picom (pretty much the only maintained one) Xcompmgr hasn't worked in the past for me and it's barely(at all?) maintained. TearFree modesetting has been in master for years but since freedesktop never does releases it's not available anywhere while it's on default on xlibre.
For now there's the guix-xlibre channel but it was rejected from guix repo because they didn't want to deal with it upstream. It's not a huge deal since channels exist you could probably use a later commit (you're gonna compile anyway since there's no build system for the guix-libre channel) with those features for yourself or package it since Guix is source-based.

The actual interesting new display server is Arcan. It's alot more than just a display server which also makes it harder to understand what it's actually supposed to do. I

>>154
other countries are working on laws similar its sadly inevitable
and outside of the context of these laws it's still useful for actual parental controls or shared Computers
also Artix is actually just a useless pile of shit and doesn't do anything to support upstream arch (this is also a problem with devuan). systemd is more than a init-system and thats why it's really useful for servers and most distros use it. it' intended for servers. i can understand not wanting to use it on desktop but acting like it's the devil because it does what its made to do well it just stupid lol???

154 better than init systems like runit or openrc

|newer thing better

again. for the 100th time. how is it better? it by your admission has shit nobody uses. do you love complexity for complexities sake? what does init need to do? start programs, give you the ability to stop them, restart them if they stop - do all this as root. how is runit not better for this? why do you need more? why can't that more be done through layers you yourself add (ie i want to run some kind of service with 5 different demons that all need to run together, throw those into a docker or qemu image and run the image as a service, holy shit i invented service grouping on runit an init that is literaly just about 100 lines of C and stupidly fast. )

>>155
if you want something that's compatible with x11 (keep in mind x11 and xorg are essentially just a protocol) look at "phoenix" my problem with Xlibre is it isn't dealing with the problem of legacy code but its doing it in a way thats more likely to break compatibility with x11 apps. the problem with x11 and why nobody wants to work on it is because xorg is built ontop of 30 years of code and operating system support. so "phoenix" exists to make a x11 compatible server from scratch (since again x11 is just a network protocol) instead of dealing with and adding onto 30 years of code

>>158
better init system? maybe? systemd is more than a initsytem though it manages logging for services network user access can it has a lot of things to protect kernel access to provide better system security/stability generating logs and actual useful errors in-case of a system crash as-well as device handling. Linux is far more used in server than on the desktop and will most likely remain this way for decades to come. and it all being under a single suite makes it a billion times easier to config and integrate together cleanly. software minimalism is BS and you'll find out quickly the short comings quickly in production

logging for services

runit has logging hooks and works with pretty much any logger. I use svlog, i can group things and search with cat + grep.

network user access

i don't understand, how does the kernel + whatever other network daemon not already handle this? is sshd not secure? sounds like a problem that sshd would be better possitioned to solve.

protect kernel access

protect it from what? protect it how? is kernel access permission, user groups, file permissions not able to solve these issues?

better stability

how? if a program is unstable, it is unstable. How does making some bloated init a middle manager poking services all the time actually help rather than just add to ram bueracracy bloat. (fix the instability of the service, that's where the theoretical problem is)

actual useful errors in-case of a system crash

how are svlog logs not "actually useful"?

device handling

the kernels literal job?

Linux is far more used in server

they are already using vm's inside of vms inside vms to manage things, because to do otherwise is too much cognative load.

if you'd rather learn 100 things useable in exactly one situation each, rather than 1 thing usable in 100 different situations, i don't know what to tell you.

>>161
yeah good luck with that i guess (it's kinda like systemd is a bunch of different tools under a suite) also you try reading my post a bit slower it might help you properly understand it

no i get that it is a bunch of different tools (that already exist and already use a common interface)
good luck rereading man pages every time you need to do something new because you can't use what you already know to do it. non-transferable skills are awesome.

i want to go full magia baiser and lock people in a room with only a plan9 grid connected terminal for comfort until the get it.

>>159
I know what an X server is.
I just want to run my Window Manager without tearing or a heavy compositor. I don't care about some proof of concept X server that will be never finished. I have two options Xlibre or Xorg compiled from source. If distributions don't want to provide Xlibre for some reason (at most the Xlibre dev is some "I don't see color" boomer, there's a Nix rice with LGBTQ+ flags on the github readme. Yes it's cringe but how many Zionists do you think there are that have contributed to Emacs for example).
All the reasons you shit on systemd for can be applied to Xorg. Xorg has tools to control a plethora of things while on Wayland it's the compositor which chooses (or doesn't) to implement these features and most Wayland compositors kinda suck or don't implement all of these features.
If I wanted to switch from EXWM to a Wayland compositor I'd use EWM which was recently bootstrapped with Claude while EXWM has been around for years and has a bunch of useful packages like exwm-edit. Also what's the security benefit from Wayland when it's all dependent on how the compositor is implemented. Are huge desktop environments like GNOME or KDE safeS Is a recent compositor made with the help of AI safe? Or do I just take the theoretical security risk and continue using my Window Manager that already works for me? The Pure-GTK version of Emacs is slower than X11 versions of Emacs ON WAYLAND! Not even accounting the difference between Emacs with alternative toolkits like Lucid which work better for remote workflows this is a point for both X11 itself and non-GTK versions of Emacs.
>>157
Why should it they explicitly mention this and ship their own repos. If you're using Arch's repos you're own your own. I'm not an Arch user or like that the base repo is smaller and you're reliant on random scripts. Nix and Guix are already easy to package for yourself and Nix is larger than the AUR. But I know there's a learning curve and some people like the traditional way of doing things. Artix actually packages alot of services for alternative init systems that you could use on other distros after some manual work and their repos contain some software not even in the Arch repos.
>>158
Rephrase: I mean there are other init systems that implement the features of systemd without all the baggage. The closest is dinit. It's great if runit works for you or you have some a list of init scripts that you've been writing yourself.

>>164
i wanna try plan9/9front i just hate using vitrual machines and nothing i have works with it from what i know adventually i wanna get like a thinkpad t420 but at the same time i don't really need it (plus im broke) i think plan9 has a risc-v port i wonder if it works on any actual boards and how hard it would be to port to something like my visionfive 2 (been rotting on my desk for like 2 years now)

>>84
wow youre an idiot

hate using vitrual machines

/why/ that seems so bizare. qemu is fantastic.

risc-v

port is a strong word. It has risc-v compiler, and has been ported to specific IoC but the ARM and Risc-V situation make it borderline impossible to make proper ports. Basically every single board needs to be bootstrapped from scratch. And there are thousands of them each with different configurations. The only reason that this isn't a problem with linux is because for all these boards the devs themselves go through that individual porting work themselves so they can have something other than a paperweight to sell.

>>168
qemu is fantastic don't get me wrong i just find using VM's for like "more regular use" instead of like testing like code really clunky. and with the risc-v stuff yeah it's def a bunch of work though vf2 has some meh documentation but it was pretty popular for a while so it had decent user porting efforts (plus working with it baremetal myself) the bootstrapping isn't too bad from my experience (though really you'll be writing drivers from scratch in most cases anyway)

with 9front in qemu, you don't ever have to interact with the vm after you install it and set it up as cpu server. you just drawterm into it, in the real plan9 way. you don't even need to manually start it if you set it up as a service (which is just a shell script with runit so it's as easy to write as the command that started it in the first place :winkyface: )
i've only set up vms for services in this manner, basically headless installs that run in the background, doing their specific things. each as their own service, so they can be easily started and stopped.

adventuresin9 had some videos about doing a risc-v port and writing drivers. it does seem doable but it's a bit much for a "first time" plan9 install.

I am using Artix on a Framework 16. Window manager is DWM on top of an xlibre server.