Ah yes, review the PKGBUILD for every AUR update. Luckily I do this while I’m rereading the ToS every time those get changed for all my software as well.
When I finish that I intend to read the changelog in git for each of the commits since the last update.
A man of culture
This is just a regular OS. Gui is bloat because DE is bloat.
Window manager is all you need, the terminal is the file explorer. Free your mind!
Unless that’s how people are designing front ends for models, it literally DOESN’T work like that. It works like that until you finish training an embedding model with masking related tasks, but that’s the tip of the iceberg. The input vector, after being tokenized, is ingested wholesale. Now there’s sometimes funny business to manage the size of a context window effectively but this isn’t that unless you’re home-rolling and you’re caching your own inputs or something before you give it to the model.
The joke is that stopping discussion is healthy (which was obviously wrong). So I said I was stopping you–and thus the discussion–and then showed it was healthy with a salad.
I get it. You can’t get by “Ai iS slOp” at top level comments anymore. I get that kind of ending because I would add it… but then I also don’t mind collecting downvotes so ymmv I guess.
I feel like there needs to be a dedicated post (and I don’t want to write it, but maybe I eventually will) that outlines what a model really is. It is not just a statistical text prediction machine unless you are being so loose with the definition of “statistical” that it doesn’t even mean anything anymore.
A decent example of a statistical text prediction machine is the middle word suggested by your phone when you’re using the keyboard. An LLM is not that.
In the most general terms, this kind of language model tokenizes a corpus of text based on a vocabulary (which is probably more than just the words in the dictionary), uses an embedding model to translate these tokens into a vector of semantic “meaning” which minimized loss in a bidirectional encoding (probably), that is then trained against a rubric for one or more topic area questions, retrained for instruction and explainability, retrained with reinforcement learning and human feedback to provide guardrails, and retrained again to make use of supplemental materials not part of the original training corpus (resource augmented generation), then distilled, then probably scaled and fine tuned against topic areas of choice (like coding or Korean or whatever) and maybe THEN made available to people to use. There are generally more parts to curriculum learning even than that but it’s a representative-ish start.
My point being that, yes, it would be nuts to pose ANY question to a predictor that says “with 84% probability, the word that is most likely follows ‘I really like’ is ‘gooning’ on reddit”, but even Grok is wildly more sophisticated than that and Grok is terrible.
Edit: And also I really like your take at the start of this thread: user error is a pretty huge problem in this space.
A little personal flourish doesn’t invalidate the rest IMO. Humans get aggravated and humans are aggravating.
20 line commit: 5 issues and suggestions.
500 line commit: “looks good to me!”
I’m just gonna stop you right there.
🥗
It sounds like you are already doing it! Fixing annoying stuff as it comes up feels painful and tedious, but that’s the good instruction you’ll remember. Hopefully you aren’t hitting too many work stoppage type issues. Definitely reach out for help for those, but be prepared for uptight people to tell you “RtFm noOB”. You’ll get it done.
My only advice is to give a little bit of time to learning at least basic bash commands (search and print a cheat sheet) and basic Vi / Vim. Vim is what got me to Linux in the first place to be honest. Those 2 things will give you the latitude to move around and work on any issue you might come across as any user even if you “break” something.
Good luck! You’re already doing the hard stuff.
Happy cake day. I was asking, before I dive into the hunt, if you saw any docs for doing the same sort of firmware update / swap for plain old irobot vacuums.
Idk how motivated to be to go do this. Just started reading about valetudo.
COA => course of action
Is this a COA for irobot Roombas?
Big if. That’s a trial and error journey unless some of the tools are part of your job description.
Edit: downvoted by Linus Torvalds I guess. Can’t even follow a video and hope it’s right unless you just duplicate a setup
Bro. How could you do that and not post to tell other people the things they like are wrong?
I carefully curate app permissions and the list of what is installed, but for me at this point the juice of degoogling or changing OS is not worth the squeeze. I’m willing to be the product a little bit to have seamless use on my phone.
When it comes to PC I am willing to take more steps.
Couldn’t say, for me it was way way easier than ESXI which was my first break into the space. And also more complete / straightforward than bare metal which was what I had been doing before unraid.
I paid for the lifetime license. No regrets.
Because it’s easy and does all the hard stuff out of the box? Also any sized drives!
Oh hi I picked up Linux for the CLI and shell and the UI for me has nothing to do with it.
There is no easy way to break into the scene and unraid is a one stop shop. So you want to set up a few little projects on your own? It’s learning containerization, learning networking and NAT, figuring out filesystems (and shares and share locations) and backup strategies, how to integrate with VPN, deployment strategies and templates (think Ansible, docker compose, make scripts, etc). There’s a shitload to know and not a “for dummies” place to learn it.
Considering the “easy” first project of ARR suite + jackett, integrate with transmission, and integrate with jellyfin or Plex: this is not a couple hours of work if you’ve never done it before. With unraid it’s probably one video tutorial and less than an hour? Idk I haven’t done that one yet. But it’s a common request.
There are a lot of things that need to hang together for a good homelab to work, and unraid for me has made it so I don’t have to spend all my time doing plumbing and background work to try a project and see if I even want to use it.
I would absolutely do a 101 on self hosting, but it seems everybody has different priorities on what to host and how so it’s probably not cut and dry to implement.
Who is having breaking update issues anymore in 2026? I’ve been running vanilla Arch for 10 years and the only times that has happened (there have been a handful I guess) the archwiki says “hey there’s a breaking change run these 2 commands” and it’s fixed. As a beginner on Linux I actually switched to Arch because every Ubuntu issue I googled was 6 to 10 lines to fix while arch was 1 to 3 lines. The only problem is that the OS expects that you be able to read, which is sometimes tough.
I can’t imagine being on a system that is multiple major releases behind on basic things like nvim and python. I guess if you’re content not to use anything remotely current it makes sense.