- Use the Daemon, it starts a new client in a fraction of a second
- Improve your config and it’ll start in under a second anyways
Emacs had some “premade IDE” project I recall that I tried and wasn’t that enthusiastic about.
Doom Emacs, spacemacs, etc.
And there are plenty of nvim “distros” like that (lazyvim for example).
They make getting started pretty easy. I’ve been using Doom for years and never bothered to make a full config of my own.
AI is quite fit for the task of understanding
Sure, and parrots are amazing at spotting fallacies like cherry picking…
Icy peepee
Or
I see peepeee
???
73 and 76, but I got them mixed up, ed is older.
That’s for original Emacs though, the gnu version came out in 85
Damn you broke my brain for a second there. I thought you meant that nixos replaced k8s, and was wondering what the hell are you talking about.
It’s the only distro I’ve dropped because the upgrade was
Worst upgrade experience ever…
Nix > distrobox for this scenario
Hell no, Emacs and nvim UX is far superior. I won’t ever go back to clicking.
I was talking about regular fedora. It’s not that you have to reboot, but you don’t get to use those updates until you do. The most obvious example is updating the kernel and its modules.
Linux almost never needs to reboot after an update
Doesn’t it often need a reboot to apply some updates?
I rember reading something along those lines then I was researching why Fedora installs some updates after a reboot. Most
And I think they rewrote a bunch of C libraries in order to have a better cross-platform compiler for C and zig. Or something along those lines
You can’t replace it.
Zig?
Julia, Clojure and Go. Are any of these good for a beginner or should I start with something else?
That totally depends on what you want to do.
Go should be easiest since it’s purposefully simplified in order to make learning it easier. There are some more difficult concepts, but the start should be easy enough. I know about go with tests, but it’s not really programming beginner friendly.
I’d avoid clojure as a beginner. It’s more for people who know java, but don’t want to write java. Common lisp and schemes are good for learning programming, but they’re not a popular group of languages and that can be a problem.
I’m going to have to come back to Nix/NixOS in a bit.
Use nix + home-manager first for sure. It’s far easier, and you can slowly get into it while making a list of bleeding edge packages.
I’ll probably wait until the official docs catch up as it appears that they are quite a bit behind
Skip them altogether when you’re starting out. I gave up on trying nix the first few times due to how bad they are. zero-to-nix.com is better for learning the basics of nix.
That and I’m not sure how I feel about a DSL for package management. I’d much rather use JSON or YAML, or even INI or TOML.
The closest you can get is home-manager with a list of packages in a json-like format. It’s really not practical to develop a declarative system without a programming language. A basic example would be variables, more advanced would be to write a wrapper that modifies the package so it automatically runs the required cli commands to use your dediated gpu and nixGL with specific packages (nvidia-run-mx nixVulkanNvidia-525.147.05 obs
for example).
It’s sort of like IaC where you’ve got terraform (dsl), pulumi (various languages), and cloudformation (json/yaml). Can you guess which one is universally despised?
Maybe if I were a LISP or Haskell guy.
Then you’d use guix and a dsl made within an actual programming language (much better approach IMO).
I don’t really care about the declarative/imperative thing, to me how many commands you “really need” is beside the point.
Caring is not required, but you need to at least understand the difference.
This is essentially the same argument as the people who say “git is not complex because you only really need checkout/commit/push, just ignore all the other commands.”
It’s really not.
Stage,commit,push,fetch,merge,etc. are all commands you need issue to git in order to manually create a desired state. You need to know what you’re doing, and what to do differently if there’s an issue.
home-manager switch
does all of it on its own. You don’t use a different cli command if something’s broken, you change the source of truth. All of the commands you might use in an imperative package manager like apt update/upgrade/install/remove are instead that one command.
Even home-manager has this warning at the very top of the page that basically tells you “you need to understand all the other commands first before you use this,” and “if your directory gets messed up you have to fix it yourself.”
It’s quite a disingenuous interpretation of “beware: home-manager uses the nix language and so gives nix language errors” and “choosing to create configuration files might overwrite the existing ones for that package”…
If you’re using a programming language, expect error messages specific to that language/compiler/interpreter/whatever. And it’s not like every other PM is using standardised error messages, you still need to learn to read them.
Config files aren’t generated randomly, you need to manually enable the configuration of each package. If someone is capable of getting to the info required to know how to configure a package, it’s reasonable to expect that they can guess that changing a config might overwrite the existing one.
These are exactly the same kinds of problems people have with git.
Do tell me how you can solve git problems without changing the git commands.
You’re essentially saying that the terraform cli has the exact same problems as the aws cli, and that’s just ridiculous. They both let you host your blog, but they do it in a completely different way and therefore have different issues.
It’s far better in theory, but in practice it’s got some massive issues:
In it’s current state it’s really only good for emacs, lisps, and some other languages like haskell.
And how is SEL less for a rich person than RHEL?