Tfw nixos will never gain enough traction

>tfw nixos will never gain enough traction

Other urls found in this thread:

bugzilla.redhat.com/show_bug.cgi?id=1141137
twitter.com/SFWRedditGifs

>retarded op
Checked.

>fishing for replies
Checked.

>anime image
Checked.

10/10 op.

I use it for scaling at work. Better and more reliable then hot glueing docker and puppet, ansible or chef together.

you tried for long enough to fit in

Only if it had ports...

GuixSD is better.

I switched from Arch to NixOS, and have had a pretty good time. In my circle 3 other people have switched as well. Its pretty convenient to share config stuff so one person doesn't have to solve every esoteric problem.

Its pretty rough around the edges though. The transition for someone who's not already used to fixing their broken crap is not great, and the provided versions of stuff I use are all ancient. Not to mention the hoops you have to go through with nix-shell to use Bundler or OpenGL. One person I know gave up on the whole os cause they couldn't get it to interop well with Opam.

Still, the first time you drop your config on a new computer `nixos-rebuild switch` and everything just works is the best.

Stay pure, nixers~

Does it even have a effortless install yet? I remember it having a major problem of dropping you off at a near unconfigured system.

You need to do manual partitioning, then run nixos-rebuild with your configuration.nix file.

one of the guys from nix took everything amazing about it and made it all even better
guix is the bright beautiful future of gnu and freedom
nothing but respect for nix though
thank you for paving the way

>made it even better
>implying it needs improvement
GNU reinvents the wheel yet again and wonders why their projects don't gain any traction

NixOS has enough packages there to effortlessly deploy pretty much any linux application or use it ask your desktop. What kind of traction do you think it needs?

I've been using for about a year on all of my systems and have had a grand old time. Nixpkgs development seems to be rocketing forward at breakneck speeds, I'd say they are a very successful project

using lisp for everything on every level is by itself an insanely huge improvement
you realize that guix was started by one of the main nix devs don't you?
they get it

GNU need to stop trying to push Scheme, It's detrimental as fuck to any project.

Stallman needs to fuck off with his hatred of CL.

Good.

I installed it and everything was just fine (except one small hiccup with drivers)

Then I find out that there are no 32 bit packages at all, and no support for it. This means I can't play the one FOSS game that I play that only has a 32 bit version (Wolfenstein Enemy Territory).

I quickly moved back to Debian. The clear benefits and strengths of the system (package management and central configuration) did not outweigh my user needs.

Haven't really looked too closely at it, but seems like it might be interesting. Useless to me, though.

If you want to "gain traction", take the current major release, designate it LTS or something, and get that shit FIPS validated.

Yes, FIPS is sometimes retarded and counter-productive. Yes, the process is ridiculously expensive and a colossal pain in the ass.

But those of us who are forced to use this shit are *desperate* for alternatives that aren't total garbage. You have no idea.

please explain why you think [a different lisp to the one \emph{i} am used to] is detrimental

common lisp is bloated and gross, the python of lisps
scheme is elegant, beautiful, intuitive, and easy to understand and use even if you've no experience programming

cl is the disgusting popular gyaru cow
scheme is the kawaii blonde loli
guile is scheme"s goth twin sister

>look up nixos
>systemd

Dropped

>I have no Lisp evolvement experience but I've read some dank memes on /dpt/

Pro-tip: Scheme is fucking useless, literally, R5RS is so "anorexic" it can't stand on its own for more than a couple of minutes but everybody had different ideas on how to "cure" it, so now you've got the least portable language that has ever existed where even bog-standard R5RS implementations aren't compatible.

Stay pleb.

Guix uses dmd

>implying

scheme as just an academic game and tool maybe
but guile is a legit language, perfectly usable
and this is what gnu is pushing
highly suggest checking it out
though, like the python people, the cl peoples"s minds seem to be contorted beyond repair

Not him, but you can pry my EL6 systems from my cold, dead fingers.

I do still hold out a small glimmer of hope for EL8, but F23 was still a pile of broken garbage.

Now known as the GNU Shepherd.

You are the kind of person that makes my development life hard. Supporting EL6 is a pain ;_;

i dont know anything about systemd, why do people hate it so much?

You blame me for running an old (but still maintained) OS, I blame you for lazily adopting "standards" that erode the software ecosystem. Dev wants new and shiny, ops wants stable and secure. And so it has been for millions of years.

That sounded better in my head, but you get the gist of it.

>guile is a legit language, perfectly usable
>this is what gnu is pushing
Yeah, it's so legit the GNU guys at Emacs came to the conclusion Elisp is will outlive it and attempts to port to Guile is wasteful as it could die at any moment (Like all the other dormant GNU shit).

People who don't know what they're talking about jump on the hate bandwagon to try and fit in.

I blame you for nothing, friend. Supporting many different platforms is a pain, and EL6 is no different.

Seems more like an issue with how crazy Emacs has gotten than Guile.
Difficulties in maintaining legacy compatibility aren't a language's fault.
Not to mention, attempting to switch the core language/api on a huge ancient project is kind of insane to begin with.

OK. You seem cool, then. Sorry I lumped you in with the assholes. :)

My devs finally tried to give me the "containers are where the industry is moving" speech last week, so I'm still a bit on edge after calling them on their bullshit....

Fuck off.

It started as one of several attempts at modernizing the init system. Being backed by Red Hat, systemd eventually won out and became widely adopted. That itself wouldn't have been a big deal, but it suffers from an extremely aggressive case of scope creep and has spread like a virus throughout the Linux ecosystem. It does so much, does it poorly, but pretty much everything now depends on it to some degree.

To be fair, it has resolved a fair number of real issues. Server administrators, though, tend to be a bit more critical of it, saying it goes against the "UNIX philosophy" (do one thing and do it well, portability, etc) that made Linux so great in the first place. And when you're coming from a dead-simple system of shell scripts, any seemingly unnecessary complexity and unpredictability is a massive negative.

It's also bug-ridden, nothing that can't be fixed, but a lot of it _won't_ be fixed because the developers have their heads entirely up their own assholes. My favorite example is this:

bugzilla.redhat.com/show_bug.cgi?id=1141137

Lennart Poettering in particular seems to revel in the hatred he has earned over the last few years. Look up his interactions with Linus Torvalds, good stuff.

Oh, but yes, there are a number of bullshit fallacious arguments both for and against it.

runit is better