Open Source Software Community - it's about ethics in Code of Conducts

  • 🔎 If you see an alert for analytics.usips.org, this is a self-hosted system I am running as a test. It is GDPR compliant and collects no invasive information.
  • Want to keep track of this thread?
    Accounts can bookmark posts, watch threads for updates, and jump back to where you stopped reading.
    Create account
Well if you don't have a fancy monitor then sure, it ain't broken so it'll work. But let's not act like X11 is gonna get proper support for things like VRR or HDR any time soon, or that XLibre will catch on with those features.
I'll be completely honest I have no clue how any of that shit works, but I'll take your word for it. As far as Wayland is concerned I like the peeps at wayland.fyi and their super suckless little toy compositors. Hevel seems cool if you have Plan9 autism.

kiwifriends should come together and make wayland2 without all the niggerliscious problems it has. i'll make the logo :)
I meaaaaan Hyprland is already le ebin chud compositor and Vaxry did do away with wlroots, perhaps he and DHH can slop up their own claudecumsuckland protocol, aka standard #16.
 
Which BSD?
beastieblacksun.webp
 
Am I being too AI brained when I say X or Wayland won't matter in a couple of years and whatever you're not satisfied with will be altered locally with a couple of prompts.
in the sense that people will use locally hosted AIs to help assist good developers to code and review faster so we get S-tier compatibility layers or in the sense that baljeet is going to @Claude to create the "leaning tower of india" code base that crashes all the time and is an unmaintainable niggerhell mess integrated into systemd?
 
Am I being too AI brained when I say X or Wayland won't matter in a couple of years and whatever you're not satisfied with will be altered locally with a couple of prompts.
an example: I had an old GTK 3 theme I wanted to update with a different color palette of my choice and also fix outstanding issues on. As it is with most custom themes these days (customizing anything of your OS somehow became frowned upon too, I am aware of that) it was in a state of disrepair as the author lost interest. I gave a local Qwen the theme in a programming harness, expressed my wishes and gave it hints how to test the themes (mostly, by looking at gtk3-widget-factory). It then autonomously figured out that in order not to bother me during computer use, it can start programs inside xfvb and check out how the theming goes by taking screenshots of the widget window with scrot and looking at them with it's vision capability, and controlling the mouse via xdotool. Not only did it adapt the theme nicely to the colors I picked, it also fixed all outstanding issues in that particular theme (mostly deprecation warnings, but also a few genuine errors) and wrote a fittingly colorized openbox theme, which I didn't even ask for.

The only complaint I had was a slightly weird looking gradient on buttons on mouse over, but that was quickly fixed. I just told the model to fix it.

I personally would have never spent the effort to fix that theme up to my specifications. I totally could have easily, I just would have felt too lazy and wouldn't have wanted to spend the time, especially getting into of how to even theme gtk3 which is a lot of effort for a one-off. AI has done a lot of work like this for me in recent months, just smoothing out the edges and clearing all the kinds of debris and technical debt you accumulate in over twenty years of linux usage and keeping tight control of everything yourself. I will get all kinds of stickers for saying this, but I think LLMs will help a lot in "democratizing" FOSS by making all these egomaniacs in various projects we post about here less important and with that less powerful. As you know, traditionally, the ability to modify and maintain open-source software has depended on having the necessary technical knowledge, time, and willingness to deal with other people's code. Inability to realistically do this by yourself even for a few projects at once has been abused to no end by certain people, as I am sure you are aware (while it's true that FOSS can be freely modified by anyone, how realistic is that, really?). This might come to an end. People will be able to realistically fork things for personal use without becoming a full-time maintainer for these particular projects. It's a good thing, things stand a chance to become a lot more decentralized again. I think a few of these FOSS trannies can feel that particular change in the air, that's why they're so staunchly anti-AI.
 
Last edited:
Am I being too AI brained when I say X or Wayland won't matter in a couple of years and whatever you're not satisfied with will be altered locally with a couple of prompts.
A couple of years is probably pretty optimistic to say that an entire display paradigm won't matter. But as others have subsequently pointed out, locally hosted AI is taking power over your machine from the hands of corpo-trannies and putting it back into the end user's control. I havn't even taken a poke at Omarchy but I suspect a huge proportion of the rage over it is due to it taking "agentic" to a current extreme - putting the ability to just change the code far more into an end user's hands than it was even 2 or so years ago. Certain people definitely don't like this.
 
A couple of years is probably pretty optimistic to say that an entire display paradigm won't matter. But as others have subsequently pointed out, locally hosted AI is taking power over your machine from the hands of corpo-trannies and putting it back into the end user's control. I havn't even taken a poke at Omarchy but I suspect a huge proportion of the rage over it is due to it taking "agentic" to a current extreme - putting the ability to just change the code far more into an end user's hands than it was even 2 or so years ago. Certain people definitely don't like this.
Omarchy is literally pure niggertrash. Being "agentic" is nothing to brag about.
Oh, do you have your digital slave do things on your computer, how innovative.
 
I havn't even taken a poke at Omarchy but I suspect a huge proportion of the rage over it is due to it taking "agentic" to a current extreme - putting the ability to just change the code far more into an end user's hands than it was even 2 or so years ago. Certain people definitely don't like this.
omarchy made me understand just how much better AI has gotten in terms of building applications. I remember vividly trying to get llama 3.1 70b to do a bare minimum character controller for godot like 2-3 years ago and it couldnt do that even though half of it was already done for it through the template.

Now through opencode i've:

Wrote an entire quickshell interface for hyprland, including full pywal16 support to theme kde applications any my terminal based off the wallpaper.

Wrote a image viewer with vulkan hardware acceleration, used that image viewer as a base to create a photo editor.

Wrote a game launcher for linux that automatically hooks goldberg steam emulator and pulls achievements from steam so i can get 90% of what I care about from steam without having to deal with steams bloat.

Wrote a local music player to completely replace spotify.

Ported minecraft legacy console edition to Linux/Opengl when it was heavily dependent on windows and directx 9, then ported it to Godot to modernize the dated engine.

All with only knowing the bare minimum of C/C++. It would have taken me a year+ to understand C fully enough to do what i've done in less than a month with AI. The future is now. I think in a few years every game and application will be wide open to being reversed engineered in a weekend, even on non frontier models. Anyone who thinks otherwise are living under a rock.
 
A couple of years is probably pretty optimistic to say that an entire display paradigm won't matter. But as others have subsequently pointed out, locally hosted AI is taking power over your machine from the hands of corpo-trannies and putting it back into the end user's control. I havn't even taken a poke at Omarchy but I suspect a huge proportion of the rage over it is due to it taking "agentic" to a current extreme - putting the ability to just change the code far more into an end user's hands than it was even 2 or so years ago. Certain people definitely don't like this.
Look bro, I'm subbed to DDH's RSS feed and wholeheartedly agree with most of his politics, but if you actually look at the issue trackers for Omarchy you will quickly realize that he has been entirely enslaved by the prompt jew. Where the AI is matters less than the fact that it is literal meth for idea guys. People that use it super extensively like DHH and other cultists become a one man human centipede that pats its own ass over replacing simple python or rust libs with a ~50000 LoC unportable assembly shit brick for a fraction of a second of a speed bump on the FUCKING SCREENSAVER that was already plenty fast. I used Omarchy last July - Oct because of its Hyprland dotfiles, and even then it already had keybinds for like 5 different AI as chromium webapps, and with the current culture of plugins and shit around it, everyone is just pumping opaque shit into eachother's mouths constantly. Omarchy is effectively reinventing MacOS from first principles, except instead of obfuscated proprietary code, it is de facto proprietary by way of 100% relying on AI for administration and customization. Do YOU read your AI vibeslop? All of it? Come on.
 
Last edited:
omarchy made me understand just how much better AI has gotten in terms of building applications. I remember vividly trying to get llama 3.1 70b to do a bare minimum character controller for godot like 2-3 years ago and it couldnt do that even though half of it was already done for it through the template.

Now through opencode i've:

Wrote an entire quickshell interface for hyprland, including full pywal16 support to theme kde applications any my terminal based off the wallpaper.

Wrote a image viewer with vulkan hardware acceleration, used that image viewer as a base to create a photo editor.

Wrote a game launcher for linux that automatically hooks goldberg steam emulator and pulls achievements from steam so i can get 90% of what I care about from steam without having to deal with steams bloat.

Wrote a local music player to completely replace spotify.

Ported minecraft legacy console edition to Linux/Opengl when it was heavily dependent on windows and directx 9, then ported it to Godot to modernize the dated engine.

All with only knowing the bare minimum of C/C++. It would have taken me a year+ to understand C fully enough to do what i've done in less than a month with AI. The future is now. I think in a few years every game and application will be wide open to being reversed engineered in a weekend, even on non frontier models. Anyone who thinks otherwise are living under a rock.
*written

Christ, I shudder to think how badly some of you guys code if you’re so sloppy with your English.
 
Last edited:
Does anyone know of any tools for distributed local LLM usage, meaning me and a couple buddies can hook up gpus together to get more vram and speed? I found mesh llm but couldn't get it to work.
 
Am I being too AI brained when I say X or Wayland won't matter in a couple of years and whatever you're not satisfied with will be altered locally with a couple of prompts.
Unless you are absolutely loaded, you won't be generating X or Wayland alternatives on a local machine ever, unless you have the skills and patience to wrangle a sub-frontier-level model into generating all the discrete parts of a graphics system yourself, and you can either stitch them together or wrangle your underpowered LLM into doing so for you. My real worry is that people will get so used to using ChatGPT and Claude over the next 10 years that Free Software development as a movement will become underfunded and scattered, and then AI companies will crank up how much their tokens cost, and it will wipe out the unorganized and under-funded Free Software projects ability to manually keep up with AI-driven attacks and keep the systems secure, and everything will just be fucked up. It seems like everybody with money and reach in tech is completely avoiding the mention of user freedom. All they talk about is source code, but if source code is free and plentiful, it's only going to be a commitment to user freedoms that keeps people organized and worried about the future.
 
My real worry is that people will get so used to using ChatGPT and Claude over the next 10 years that Free Software development as a movement will become underfunded and scattered, and then AI companies will crank up how much their tokens cost, and it will wipe out the unorganized and under-funded Free Software projects ability to manually keep up with AI-driven attacks and keep the systems secure, and everything will just be fucked up.

The big LLM companies are in a bit of a squeeze. Their products are becoming commoditized and interchangeable to the point that there's not much of a difference between Claude and Grok and OpenAI at least in terms of end result, so they are in competition with each other to provide the lowest prices to the consumers, subsidized by venture capital of course - for now. Meanwhile, as consumer-level hardware and open models improve, there's going to be a long tail of consumer-level users who are going to find that running Qwen on their laptop is good enough and cancel their $20-40/month subscriptions, and eventually corpo clients who find that buying a $10,000 local token-server box loaded with GLM to serve the office is a better deal than a $1000/month corpo subscription. That may be why so many execs of the big companies are already cozying up to politicians and lobbying for more government legislation for their own product, which will of course affect their smaller competitors more than them - I'm sure they'd love to get in Trump's ear to the extent that he signs an EO banning "import" of the free Chinese models, for example.

Yes, the majority of the above is speculation that I've pulled out of my ass but it sounds plausible, right? I guess we'll see.

To bring it back to FOSS development, yes, maybe the big companies will eventually either die out or rationalize their prices to the extent that FOSS developers can't use them anymore, but I think at that point the free models and consumer hardware will be sufficient to pick up the slack at at least the same level the big boys can do it now. Unfortunately we're gonna be stuck with slopcode forever.
 
Last edited:
*written

Christ, I shudder to think how badly some of you guys code if you’re so sloppy with your English.
Holy shit a grammar nazi on green reddit. Sorry for not taking a forum seriously enough to care about proper grammar. Take your PFP seriously and go back to Reddit.

Does anyone know of any tools for distributed local LLM usage, meaning me and a couple buddies can hook up gpus together to get more vram and speed? I found mesh llm but couldn't get it to work.
I don't think there are any, even if you do get them running unless your running 10gb networking to each computer bare minimum the speed is going to be atrocious once a model needs to flow over to another computer.
 
That may be why so many execs of the big companies are already cozying up to politicians and lobbying for more government legislation for their own product, which will of course affect their smaller competitors more than them - I'm sure they'd love to get in Trump's ear to the extent that he signs an EO banning "import" of the free Chinese models, for example.

Yes, the majority of the above is speculation that I've pulled out of my ass but it sounds plausible, right? I guess we'll see.
They're already trying to pull the ladder up behind them, that's what all the "AI is going to end the world!" doomposting articles are for.
 
They're already trying to pull the ladder up behind them, that's what all the "AI is going to end the world!" doomposting articles are for.
I read that as them trying to soften the blow of the eventual free money flow from investors drying up. Then they can't be blamed for their models plateauing (as they already have), because oh, its a safety concern, goy! Haven't you ever watched the Terminator o algo? Look, look, we put these two bots in a boxing ring and had them smack each other, the singularity is already upon us, better bail us out er... I mean, apply safety guardrails before it is too late!
 
but if you actually look at the issue trackers for Omarchy you will quickly realize that he has been entirely enslaved by the prompt jew. Where the AI is matters less than the fact that it is literal meth for idea guys. People that use it super extensively like DHH and other cultists become a one man human centipede
If you're out of the loop on various 'AI people' you perhaps aren't aware about both how frequent and severe this is becoming. We're talking full personality breaks where if the person in question weren't so wealthy they'd be involuntarily institutionalised. A great example is Steve Yegge who by most accounts was a productive member at both Amazon and Google (and made enough there to retire in his 40s), now wastes his time on endless AI self-referential nonsense such as gas town.

BE AWARE THIS IS BARELY 10% OF THE ABOVE LINK

The seven Gas Town roles all work together to help keep Gas Town running. And it needs your help sometimes, too; Gas Town runs on equal parts guzzoline and elbow grease.

Here are the key players and concepts:

🏙️The Town: This is your HQ. Mine is ~/gt, and all my project rigs go beneath it: gastown, beads, wyvern, efrit, etc.. The town (Go binary gt) manages and orchestrates all the workers across all your rigs. You keep it in a separate repo, mostly for the configuration.

🏗️Rigs: Each project (git repo) you put under Gas Town management is called a Rig. Some roles (Witness, Polecats, Refinery, Crew) are per-rig, while others (Mayor, Deacon, Dogs) are town-level roles. gt rig add and related commands manage your rig within the Gas Town harness. Rigs are easy to add and remove.

👤The Overseer: That’s you, Human. The eighth role. I gave you some eye paint in the picture. As the Overseer, you have an identity in the system, and your own inbox, and you can send and receive town mail. You’re the boss, the head honcho, the big cheese.

🎩The Mayor: This is the main agent you talk to most of the time. It’s your concierge and chief-of-staff. But if the Mayor is busy, all the other workers are also Claude Code, so they are all very smart and helpful. The Mayor typically kicks off most of your work convoys, and receives notifications when they finish.

😺Polecats: Gas Town is a work-swarming engine. Polecats are ephemeral per-rig workers that spin up on demand. Polecats work, often in swarms, to produce Merge Requests (MRs), then hand them off to the Merge Queue (MQ). After the merge they are fully decommissioned, though their names are recycled.

🏭Refinery: As soon as you start swarming workers, you run into the Merge Queue (MQ) problem. Your workers get into a monkey knife fight over rebasing/merging and it can get ugly. The baseline can change so much during a swarm that the final workers getting merged are trying to merge against an unrecognizable new head. They may need to completely reimagine their changes and reimplement them. This is the job of the Refinery: the engineer agent responsible for intelligently merging all changes, one at a time, to main. No work can be lost, though it is allowed to escalate.

🦉The Witness: Once you spin up enough polecats, you realize you need an agent just to watch over them and help them get un-stuck. Gas Town’s propulsion (GUPP) is effective, but still a bit flaky right now, and sometimes you will need to go hustle the polecats to get their MRs submitted, and then hustle the Refinery to deal with them. The Witness patrol helps smooth this out so it’s almost perfect for most runs.

🐺The Deacon: The deacon is the daemon beacon. It’s named for a Dennis Hopper character from Waterworld that was inspired by the character Lord Humungus in the Mad Max universe, making it a crossover. The Deacon is a Patrol Agent: it runs a “patrol” (a well-defined workflow) in a loop. Gas Town has a daemon that pings the Deacon every couple minutes and says, “Do your job.” The Deacon intelligently propagates this DYFJ signal downward to the other town workers, ensuring Gas Town stays working.

🐶Dogs: Inspired by Mick Herron’s MI5 “Dogs”, this is the Deacon’s personal crew. Unlike polecats, Dogs are town-level workers. They do things like maintenance (cleaning up stale branches, etc.) and occasional handyman work for the Deacon, such as running plugins. The Deacon’s patrol got so overloaded with responsibilities that it needed helpers, so I added the Dogs. This keeps the Deacon focused completing on its patrol, rather than getting bogged down and stuck on one of the steps. The Deacon slings work to the Dogs and they handle the grungy details.

🐕Boot the Dog: There is a special Dog named Boot who is awakened every 5 minutes by the daemon, just to check on the Deacon. That’s its only job. Boot exists because the daemon kept interrupting the Deacon with annoying heartbeats and pep talks, so now the dog gets to hear it. Boot decides if the Deacon needs a heartbeat, a nudge, a restart, or simply to be left alone, then goes back to sleep.

👷The Crew: The Crew, despite being last in the list, are the agents you’ll personally use the most, after the Mayor. The crew are per-Rig coding agents who work for the Overseer (you), and are not managed by the Witness. You choose their names and they have long-lived identities. You can spin up as many as you like. The tmux bindings let you cycle through the crew in a loop for each rig with C-b n/p. The Crew are the direct replacements for whatever workflow you used to be using. It’s just a bunch of named claude code instances that can get mail and can sling work around. The crew are great for stuff like design work, where there is a lot of back-and-forth. They’re great. You’ll love your crew.

📬Mail and Messaging

Beads are the atomic unit of work in Gas Town. A bead is a special kind of issue-tracker issue, with an ID, description, status, assignee, and so on. Beads are stored in JSON (one issue per line) and tracked in Git along with your project repo. Town mail and messaging (events) use Beads, as do other types of orchestration.

Gas Town has a two-level Beads structure: Rig beads, and Town beads.

And if the above was confusing, don't worry, there's a handy series of infographics to explain it all

gt1.webp gt2.webp gt3.webp gt4.webp gt5.webp gt6.webp gt7.webp gt8.webp

Well it's open source at least: https://github.com/gastownhall/gastown . 7700 commits in the 220 days of active development. It has been officially abandoned with 370 open issues in favour of his new closed source "software factory" Wheelhouse, which I'm sure will actually produce something and not just be an ouroboros of token burning make-work.

Bonus:
expensive said:
My solution has been to create a token tap on $200 Max accounts, which for me work out to ~30x the list-price equivalent. So in reality I'm only spending about $2800/month out of pocket for my $87k "worth" of tokens. Though that number keeps growing alarmingly.
 
If you're out of the loop on various 'AI people' you perhaps aren't aware about both how frequent and severe this is becoming. We're talking full personality breaks where if the person in question weren't so wealthy they'd be involuntarily institutionalised. A great example is Steve Yegge who by most accounts was a productive member at both Amazon and Google (and made enough there to retire in his 40s), now wastes his time on endless AI self-referential nonsense such as gas town.
are you absolutely certain that he wasn't insane from the start?
1790758354424.png
 
Back
Top Bottom