What do you think about a "SteamBSD" machine ?

Zare said:
Unless a programmer you can't know what it feels like to maintain a project over time. Even half of real programmers don't understand that because they never worked in proper software engineering but hit and run startup projects.
Click to expand...
Problem with vibe development is that everybody can do it. There's no way to create something with real value Everything is made of public ready-to-use fragments. What those developers should wonder is why there still are no products created with some kind of AI-aid that are clearly better and worth more than existing things developed by people.
 
Its not my "fault" if no one is interested to talk about the prerequisites for starting
Click to expand...

If you dont have a complete knowledge of how the Kernel,
Linuxulator and graphics stack work then what prompt are you going to give Claude.

You won't to able to give Claude any technical directions to follow,
just a generic prompt like "make steam work on freebsd"

So your stuck before you have begun and are asking what are the "prerequisites for starting"
and want people to give you prompts for Claude because you dont have the technical knowledge of the subject,
and dont know what questions to ask Claude or what problem you are trying to solve.

Programming is not act of spitting out code.
Programming happens in the head. The code is a formalization of programmer's mindset.
Click to expand...

Exactly you need to have an in depth knowledge of how something works,
and that only comes with reading all the documentation, man pages and understanding the language you are programming in.

Based on that knowledge you then need to logically figure out the solution.
 
Zare said:
This is the biggest mistake. Programming is not act of spitting out code. Programming happens in the head. The code is a formalization of programmer's mindset. Nothing apart from living beings has mind. Lets not go into metaphysics mumbo jumbo. We live, computers don't.

A medium sized project built by layman vibe code is this

View attachment 27341

Imagine yourself as a proper architect and somebody comes to you, bro, why wouldn't you work on my home, I want to collaborate.
That is the answer why vibe projects struggle to get real programmers on board. You don't actually know what kind of monstrosity AI created for you.

Unless a programmer you can't know what it feels like to maintain a project over time. Even half of real programmers don't understand that because they never worked in proper software engineering but hit and run startup projects.
Click to expand...

You know,I feel to be a programmer even if I don't know how to write code in programming languages, just because I am the director of the works. Claude writes the code. And since I have successfully completed several complex projects which work great, I appear to be a programmer without actually knowing any programming languages. Why? Because I make the right calls regarding the choices Claude presents to me—choices framed in simple terms. This allows me to grasp the essence of what will be done, even without knowing the specific syntax or idioms. The feeling that few people here know how "vibe coding" really works is resurfacing.
 
MG said:
Problem with vibe development is that everybody can do it.
Click to expand...

That's not true. I want to explain why. A few days ago, I suggested to a surveyor friend of mine to use Claude to handle some complex, tedious tasks related to his work. Yet, when I actually sat down with the prompt interface, I didn't know what to tell Claude to get it to set up the project he had in mind. That’s because I’m not a surveyor; I lack the technical vocabulary needed to craft prompts specific to that field. I do, however, have the expertise of a Linux and Unix systems administrator, having used those systems since the 90s. So, the answer is no: without a solid background in a specific field, you simply cannot formulate the right prompts for Claude. Knowing how to use AI doesn't mean you can do absolutely anything you can imagine. You can write generic prompts if you really insists but doing so would leave the AI with too much leeway; it would likely produce results you didn't want, simply because you lacked an adequate level of depth of knowledge required for that particular area of expertise.
 
ZioMario said:
You know,I feel to be a programmer even if I don't know how to write code in programming languages, just because I am the director of the works. Claude writes the code. And since I have successfully completed several complex projects which work great, I appear to be a programmer without actually knowing any programming languages. Why? Because I make the right calls regarding the choices Claude presents to me—choices framed in simple terms. This allows me to grasp the essence of what will be done, even without knowing the specific syntax or idioms. The feeling that few people here know how "vibe coding" really works is resurfacing.
Click to expand...

It's not your work. Nor do you undestand the logic that agent dumped onto your screen. It's synthetic slop. If that slop ever breaks, or gets exploited, or becomes unmaintainable; who's going to fix it? You? No. It will become rotten abandonware. A waste of bandwidth and energy (and precious water cooling these f***king DCs) in the end.

If you went to a comedy show, and the comedian played a recording of jokes instead of entertaining you - would you feel shortchanged, cheated out of your money. You don't cheat with code. Learn how to code and do it correctly.

Why not just work on porting more games to FreeBSD? WINE contributions would be more appreciated.
 
Beastie7 said:
It's not your work. Nor do you undestand the logic that agent dumped onto your screen. It's synthetic slop. If that slop ever breaks, or gets exploited, or becomes unmaintainable; who's going to fix it? You? No. It will become rotten abandonware. A waste of bandwidth and energy (and precious water cooling these f***king DCs) in the end.

If you went to a comedy show, and the comedian played a recording of jokes instead of entertaining you - would you feel shortchanged, cheated out of your money. You don't cheat with code. Learn how to code and do it correctly.

Why not just work on porting more games to FreeBSD? WINE contributions would be more appreciated.
Click to expand...

You don't know me. You don't know what I know or what I don't know. So, I want to invite you to focus on what you can do or what you do not know, and to exercise greater caution when speaking about the skills of the other users. Thank you.
 
You know,I feel to be a programmer even if I don't know how to write code in programming languages, just because I am the director of the works.
Click to expand...

Thats like telling a builder build me a 6 foot high wall for my garden,
and then claiming your a builder.

Your neighbour comes round and says i hear your a builder,
thats a nice wall you built can you build me one as well.

How many bricks and how much cement will you need to build me a 6 foot high wall for my garden,
you reply "i dont know, i cant build a wall, im a vibe builder"

Neighbour goes away very confused.
 
ZioMario said:
You don't know me. You don't know what I know or what I don't know. So, I want to invite you to focus on what you can do or what you do not know, and to exercise greater caution when speaking about the skills of the other users. Thank you.
Click to expand...

You said it yourself. And the creation of this thread outlines that. Taking someone else's work and throwing slop on it isn't an honorable thing to do either. That "Mr Know it all" you quoted was responsible for Steam even running on FreeBSD to begin with.

Good luck though.
 
NapoleonWils0n said:
Thats like telling a builder build me a 6 foot high wall for my garden,and then claiming your a builder.
Click to expand...

Your comparison doesn't work because you're using old-fashioned thought patterns. To work, you'd have to assume the engineer guides me step by step, answers every question I ask to help me understand what he's doing and why, and then asks if I'm okay with it. You've omitted all that to make your example fits well. Times have changed, and perhaps you still don't accept it. Today, the engineer shares his knowledge with students or between other engineers.

So. I don't feel myself like a builder, but like a co-builder. The users and the AI work together, each (or every ?) one with specific roles.
 
Beastie7 said:
You said it yourself. And the creation of this thread outlines that. Taking someone else's work and throwing slop on it isn't an honorable thing to do either. That "Mr Know it all" you quoted was responsible for Steam even running on FreeBSD to begin with.

Good luck though.
Click to expand...

I called him "Mr Know it all" because he called Maturin "Mr. Smartass". Without questioning his skills, but only his particular way of expressing himself. I have specified that the tone was ironic,maybe you didn't read it.
 
ZioMario said:
That's not true. I want to explain why. A few days ago, I suggested to a surveyor friend of mine to use Claude to handle some complex, tedious tasks related to his work. Yet, when I actually sat down with the prompt interface, I didn't know what to tell Claude to get it to set up the project he had in mind. That’s because I’m not a surveyor; I lack the technical vocabulary needed to craft prompts specific to that field. I do, however, have the expertise of a Linux and Unix systems administrator, having used those systems since the 90s. So, the answer is no: without a solid background in a specific field, you simply cannot formulate the right prompts for Claude. Knowing how to use AI doesn't mean you can do absolutely anything you can imagine. You can write generic prompts if you really insists but doing so would leave the AI with too much leeway; it would likely produce results you didn't want, simply because you lacked the depth of knowledge required for that particular area of expertise.
Click to expand...
If it was true, the established market of software should be replaced with AI-systems by now. But AI can't even point to errors in code better than a compiler, let alone solve anything. It can repeat the substance of programming books and that's it. It's not approveable constructive like a person.
 
MG said:
If it was true, the established market of software should be replaced with AI-systems by now. But AI can't even point to errors in code better than a compiler, let alone solve anything. It can repeat hte substance of programming books and that's it. It's not approveable constructive like a person.
Click to expand...

And it doesn't have to be. AI is a productivity accelerator. It's the driver who has to put his mind behind it. Once the driver has established a course, they can let the vehicle go. Except for correcting it every now and then, when he realizes that it's doing unwanted things.
 
Taking someone else's work and throwing slop on it isn't an honorable thing to do either. That "Mr Know it all" you quoted was responsible for Steam even running on FreeBSD to begin with.
Click to expand...

Probably not the best move if your asking if he wants to collaborate,
even if it was ironic.

shkhln doesnnt like any of my projects that use the dummy-uvm.so,
like getting Davinci Resolve working with the Linuxulator, a Rocky Linux Jail or Podman.

Using Podman to get Comfy-ui, demux, whisperx, ollama and blender working with Cuda
and ffmpeg working with nvenc.

Which is absolutely fine,
i have no issues with that at all.

And those projects arent even vibe coded.

They are all my own work the result of reading massive amounts of documentation,
having a in depth knowledge of multiple subjects and logically thinking things through.

So i would guess he even less keen on people creating vibe coded projects using his work.
 
MG said:
If it was true, the established market of software should be replaced with AI-systems by now. But AI can't even point to errors in code better than a compiler, let alone solve anything. It can repeat hte substance of programming books and that's it. It's not approveable constructive like a person.
Click to expand...

I'm not sure why people think they can become "AI driver programmers" but they cannot become "AI driver lawyers" or "AI driver doctors"

ZioMario said:
And it doesn't have to be. AI is a productivity accelerator. It's the driver who has to put his mind behind it. Once the driver has established a course, they can let the vehicle go. Except for correcting it every now and then, when he realizes that it's doing unwanted things.
Click to expand...

Ok then drive. Don't ask for help, collaboration, or paste your logs of LLM discussions in hope somebody throws you a comment. Do you think you can establish a course, sit back and steer? Then do it. There are 4 pages of talk here about vapourware.

There is a list of games running on the current LSU. Some of them are problematic, have no sound. I also believe joystick support is quite flaky (in emulated Steam overall) but don't quote me on that. A lot of games simply don't run and we're not talking about anticheat ones. We can run only a percentage of what a Linux distro can. Why don't you orchestrate AI to fix some of it, understand what it did, understand whether it has implications on the rest (we cannot tell to AI to test all currently supported games to catch regressions), in order so you can own, defend, and discuss the change with project owners who you ask to accept your patch.

The PC box and the rest are just something that emerges once LSU starts catching up to Steam Linux state.
 
>And those projects arent even vibe coded projects are
>They are all my own work
Click to expand...

Vibe-coded projects belong to those who create them (or even to those who create them) in the same way that what you create belongs to you (or even to you) because an AI is a tool, not a person.
 
Zare said:
I'm not sure why people think they can become "AI driver programmers" but they cannot become "AI driver lawyers" or "AI driver doctors"

Ok then drive. Don't ask for help, collaboration, or paste your logs of LLM discussions in hope somebody throws you a comment. Do you think you can establish a course, sit back and steer? Then do it. There are 4 pages of talk here about vapourware.

There is a list of games running on the current LSU. Some of them are problematic, have no sound. I also believe joystick support is quite flaky (in emulated Steam overall) but don't quote me on that. A lot of games simply don't run and we're not talking about anticheat ones. We can run only a percentage of what a Linux distro can. Why don't you orchestrate AI to fix some of it, understand what it did, understand whether it has implications on the rest (we cannot tell to AI to test all currently supported games to catch regressions), in order so you can own, defend, and discuss the change with project owners who you ask to accept your patch.

The PC box and the rest are just something that emerges once LSU starts catching up to Steam Linux state.
Click to expand...

This is the off topic section,not a programming section. And I'm still considering whether and how many obstacles the project will encounter. Because embarking on a complicated project and doing it for yourself is one thing, doing it for yourself AND for others is another. And I don't see anything wrong with seeking out people interested in collaborating. Why shouldn't I come here and share with you what I'd like to do ?
 
Zare said:
Of course. But you need to take comments and criticism too.
Click to expand...

Of course. But it's tough when you're faced with a wall of intense criticism that tests your will so intensely. And very few encouragements or positive criticism for the advantages obtained using AI.
 
And I don't see anything wrong with seeking out people interested in collaborating.
Click to expand...

How are people going to collaborate.

They wont be able to have a technical discussion about the kernel,
linuxulator, graphics stack or programming with you because thats not something you know about.

Are you expecting people to have a three way conversation with you and Claude,
were they give you a prompt you ask claude and then paste the response back into the forum.

How else is it going to work
 
ZioMario said:
32-bit, and this is the hard part. The Steam client is i386, and the WoW64 situation on FreeBSD is in bad shape: on the forums in June 2026 you can read that on 15.1 wine-devel only ran with WINEARCH win64/wow64 and the syswow64 folder stayed empty, and that getting a 32-bit game working with DXVK required installing wine from regular pkg plus pkg32. So the one thing you cannot route around — the client — is precisely the one sitting on the most fragile path. This needs to be measured first.


Claude.
Click to expand...
As expected of someone using an LLM, your post is from August 18th, wine-devel landed new wow64 on July 30th, 2026. And wine-proton landed new wow64 support in July 14th, 2026.
Both project had patches for it waiting for at least a years ago, so maybe stop spreading misinformation ?
 
monwarez said:
As expected of someone using an LLM, your post is from August 18th, wine-devel landed new wow64 on July 30th, 2026. And wine-proton landed new wow64 support in July 14th, 2026.
Both project had patches for it waiting for at least a years ago, so maybe stop spreading misinformation ?
Click to expand...

Stop with this bad faith behavior addressed to me and/or AI. Anyone could have made the same mistake. Forums are designed to stay informed, not to criticize constantly using harsh and unsociable tones. I'm tired of the constant criticism that mostly doesn't offer useful solutions. Just correct the inaccuracy and stop adding fuel to the fire of the constant, non-constructive criticism I receive for using AI. Someone must intervene to stop this drip-feed. I hope you understand that my outburst is not directed specifically at/to you.
 
How telling truth is harsh, if you are spreading misinformation, people need to know. You didn't do the proper research, I did discussed new wow64 on this very forums, see right here:
monwarez

general/other Post in thread '[wine-devel] [FreeBSD 15.0] No update for 32-bit Wine'

siealex said:
Compat14x is already installed (why?). Then I downloaded the 32-bit package from the 14.x repo mirror and executed the last command from /usr/local/share/wine/pkg32.sh (as it is displayed when running the script), but with "install" and the local file path at the end. Is this the only way to make it run on 15.0?
Click to expand...
See https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=291328
You have to force it to use the 14.3 repos, which require to change the current pkg32.sh scripts.

The plan was to have all the 15 release cycle times to properly port and test new wow64. As of now, only...

I don't see why one should stop criticizing of anyone LLM usage, especially when we well know that it will have a high impact on the planet, that it basically is a process to just steal everyone code and put it on another license.
It is not because the source code is open source that it is a free game, there is always license terms and authorship. LLM tools cannot guarantee that those terms are always respected, and you can't really prove it.

Also it is kinda ironic that you want to use the same technologies that is the cause of inflated price in the ram and storage market, and thus graphics card, making this project even more expensive.

To stay on the topics, this is likely already working on FreeBSD.
Given you will probably feed this to the AI machine I won't say more.
 
fahrenheit said:
Well, the steamdeck hardware itself is well supported. The steam machine hardware also (both are nothing fancy). So, hardware support is not the issue.
From using an integrated radeon 780m, I know the weak spot is software. 6.12 DRM support is what you have at the moment, it doesn't matter that you use a compat/linux layer with upgraded support (rl9, cachyos, etc), that's the limit you get. While we do have proton and wine support both have issues specific for FreeBSD (lack of esync, ntsync, no cgroups, etc, poor support for 32bit).
Then you get to the stuff that makes steamos, on at least steamdeck, great, the level of integration with hardware (i.e. working si0x).

So, lofty goals, but maybe start small and get what support we do have to work correctly first: no rebar support on rdna4, xe graphics support, make guc firmware loading work with suspend, fix the lack of native vulkan support on compat/linux, fix the kmod repos, etc. Stuff that would benefit everyone instead of just "gamers".
Click to expand...
Btw we do have ntsync landed in CURRENT, and in 15-STABLE, and finally available in wine-devel.
As for 32bit support it is solved by new wow64, and I have yet to get any report of it not working.
 
Anyway,can you explain how an AI can keeps itself informed about what works and what doesn't, if websites protect themselves against its data-gathering attempts using Anubis ?
 
monwarez said:
How telling truth is harsh, if you are spreading misinformation, people need to know. You didn't do the proper research, I did discussed new wow64 on this very forums, see right here:
monwarez

general/other Post in thread '[wine-devel] [FreeBSD 15.0] No update for 32-bit Wine'

siealex said:
Compat14x is already installed (why?). Then I downloaded the 32-bit package from the 14.x repo mirror and executed the last command from /usr/local/share/wine/pkg32.sh (as it is displayed when running the script), but with "install" and the local file path at the end. Is this the only way to make it run on 15.0?
Click to expand...
See https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=291328
You have to force it to use the 14.3 repos, which require to change the current pkg32.sh scripts.

The plan was to have all the 15 release cycle times to properly port and test new wow64. As of now, only...

I don't see why one should stop criticizing of anyone LLM usage, especially when we well know that it will have a high impact on the planet, that it basically is a process to just steal everyone code and put it on another license.
It is not because the source code is open source that it is a free game, there is always license terms and authorship. LLM tools cannot guarantee that those terms are always respected, and you can't really prove it.

Also it is kinda ironic that you want to use the same technologies that is the cause of inflated price in the ram and storage market, and thus graphics card, making this project even more expensive.

To stay on the topics, this is likely already working on FreeBSD.
Given you will probably feed this to the AI machine I won't say more.
Click to expand...

In my opinion, your criticisms are indiscriminate—they target AI and users alike. They ought to be directed specifically at those users who employ AI irresponsibly. How many actually act that way ? Do you have this information ? We cannot foster a constructive, socially useful dialogue by focusing mostly on what is wrong—or rather, on what you (or we) *think* is wrong—while launching indiscriminate judgments; that is not the way to make progress. Let’s set aside the actions of the irresponsible few and focus instead on what we, as responsible individuals, can do. What an AI does doesn't matter, because an AI is a passive tool. What matters is how we use it—what *we* do.
 
You must log in or register to reply here.
Back
Top