Rust wgpu now has a
-
Rust wgpu now has a .agents/skills, a .claude/skills, an AGENTS.md, and a CLAUDE.md.
Increasingly wondering if my goal of living in a way that I never benefit from the actions of generative AI users and generative AI users never benefit from the actions of me is going to inherently mean never using hardware-accelerated video on a computer again.
@mcc@mastodon.social this is most likely true, but even so, the firmware of your computer is being slop coded by random corporate engineers
I think to truly adhere to this you'd require open hardware and open software for that hardware which also isn't GenAI enabled
to be entirely honest I've been thinking about this a lot and I have no solutions -
Anyway, I spent the last week writing documentation for Redox. The Redox project seems to be happy for my documentation patches. One thimble full of water, out of the sinking boat.
Thank you for speaking about this. I don't have anything interesting or useful to add, but I appreciate not feeling alone in my desire to avoid slop.
I hope you're successful, no matter which route you ultimately take.
I'm also encouraged to learn that both Servo and RedoxOS aren't accepting LLM contributions. I've been frustrated that - what appears to be - so much of the Rust community including the core Rust language itself has embraced LLMs.
-
What I mean is, my daydream GPU was going to be designed to be good for the kinds of 2D acceleration you need to like, run a web browser, scroll a window, move a window around. Want to do anything 3D? It's slow or it kicks you back out to MESA software rendering. I just wanted a computer I can use in a day to day way that felt like it was mine and I owned it.
There are a few rudimentary open source GPUs, but none with this focus, and critically, *none I know that support Vulkan*.
@mcc I wonder the chip in some snes game would be suitable for such usage
-
@mcc wgpu is a slightly simpler API to use, but I'm not sure I'd categorize it as easier to implement. Basically it means doing a whole bunch of extra work in the driver. Whereas vulkan itself has a bunch of shared implementation in e.g. Mesa, and a huge test suite, and also much of it is optional anyway.
@dotstdy The problem I hit to begin with was that the open source community is no longer trustworthy and a well-known, well-understood, well-tested system may sometime after September 2025 have been compromised with "AI code assistant" slop. Can I be certain Mesa is not, itself, slopped? Can I be certain it will continue to not be slopped?
-
@mcc instead of going with an fpga have you looked at some of the simpler GPU cores (like the powervr ones)?
@mcc rockchip makes SoCs that are suitable to use in a laptop that have powervr cores
-
@mcc@mastodon.social this is most likely true, but even so, the firmware of your computer is being slop coded by random corporate engineers
I think to truly adhere to this you'd require open hardware and open software for that hardware which also isn't GenAI enabled
to be entirely honest I've been thinking about this a lot and I have no solutions@froge About half of the rest of the thread is about open hardware.
I don't know if my current computer can run LibreBoot or not. I don't know for an absolute fact what the LibreBoot maintainer's values regarding "AI code assistants" is.
-
@mcc instead of going with an fpga have you looked at some of the simpler GPU cores (like the powervr ones)?
@bob What got me into this problem in the first place was that a product produced by a corporation may be discontinued, or be replaced with something that requires unacceptable compromises. PowerVR may go out of business, be bought by AMD, or become an AI company like the rest of them. If the goal was to just keep old stuff working and running, I could just stay on Windows 10 forever, keep binary-hacking it as new CVEs are found, and then I don't to put up with the many downsides of Linux.
-
This thread is probably meandering and maybe solipsistic. My point is this: My weird background means I can look at this staircase full of missing stairs, identify one step, and go, "I can fix that". But I don't know if there's enough of me to fix multiple steps. I haven't decided what to do from here. Pick a step and fix it anyway? Pick a "fun" project, like the GPU [which face it, even if I finished it chances are not high more than 3 people would ever use it]? Give up and learn electric bass?
@mcc I would recommend that you learn electric bass regardless, because it is such a wonderful instrument.
-
Thank you for speaking about this. I don't have anything interesting or useful to add, but I appreciate not feeling alone in my desire to avoid slop.
I hope you're successful, no matter which route you ultimately take.
I'm also encouraged to learn that both Servo and RedoxOS aren't accepting LLM contributions. I've been frustrated that - what appears to be - so much of the Rust community including the core Rust language itself has embraced LLMs.
@firebreathingduck Yes. Which creates a real problem if you want to write a slopfree program in Rust. Suddenly Rust's giant library ecosystem goes from being the core benefit of the language to being a giant source of risk.
-
@mcc I wonder the chip in some snes game would be suitable for such usage
@gkrnours uhh…
saturn, maybe.
-
@mcc rockchip makes SoCs that are suitable to use in a laptop that have powervr cores
@bob If I had the rest of the stack, then this would be a good solution in terms of "everything's deslopped except my AMD GPU; I'll just switch to the rockchip to fix that part". But since the stack is full of holes, it feels more productive to fix a hole further up the stack than try to make a project out of reorienting myself around a random Chinese company's that might get rugpulled like my last three solutions were.
-
@mcc In 2043 mcc creates the last truly slop-free, hardware-accelerated video game using Vulkan SC on a retired 737 Cockpit Display System. Worldwide only four systems remain capable of running the game as intended after the FAA and EASA finally embraced Generative AI and the population learned to accept the catastrophic loss of %0.018 of flights.
-
@mcc for completeness, i already had a PoC of this btw that i'm planning to revive https://mntre.com/media/reform_md/2022-09-29-rkx7-showcase.html
@mntmn This exact page three-four years ago is literally what started me thinking about the linux-integrated FPGA GPU lol
-
@crzwdjk That's some good clear thinking, thank you.
I am not sure but I think the bridge-to-GPU layer in Linux goes through a middle step named "DRI" or "DRM", and I think Redox plans to adopt this wholesale. I don't know to what extent DRI/DRM is about just *configuring* displays and to what extent if you've implemented the DRI/DRM layer then Mesa just does the rest for free.
@mcc DRI/DRM is the kernel part which handles confiuring displays, but also allocating memory for the GPU and managing the submission of command buffers to the GPU. The commands in the command buffers are produced by Mesa though.
-
@mcc DRI/DRM is the kernel part which handles confiuring displays, but also allocating memory for the GPU and managing the submission of command buffers to the GPU. The commands in the command buffers are produced by Mesa though.
@crzwdjk Thanks for these explanations.
-
@mcc In 2043 mcc creates the last truly slop-free, hardware-accelerated video game using Vulkan SC on a retired 737 Cockpit Display System. Worldwide only four systems remain capable of running the game as intended after the FAA and EASA finally embraced Generative AI and the population learned to accept the catastrophic loss of %0.018 of flights.
@ptrchas3 @mcc I have much hope in the gamebub [https://www.crowdsupply.com/second-bedroom/game-bub]
-
@mntmn This exact page three-four years ago is literally what started me thinking about the linux-integrated FPGA GPU lol
@mcc oh that's awesome! back then linux wasn't yet sloppified too, simpler times
next thing i'd like to try on this is haiku. also, there's a lot more performant (and OoO) risc-v soft cores now -
@bob What got me into this problem in the first place was that a product produced by a corporation may be discontinued, or be replaced with something that requires unacceptable compromises. PowerVR may go out of business, be bought by AMD, or become an AI company like the rest of them. If the goal was to just keep old stuff working and running, I could just stay on Windows 10 forever, keep binary-hacking it as new CVEs are found, and then I don't to put up with the many downsides of Linux.
@mcc but your fpga gpu is tied to a particular fpga (unless you make changes) so then you're tied to amd/xilinx or altera or lattice
-
@froge About half of the rest of the thread is about open hardware.
I don't know if my current computer can run LibreBoot or not. I don't know for an absolute fact what the LibreBoot maintainer's values regarding "AI code assistants" is.
@mcc@mastodon.social yeah I realized, I need to get better about reading the entire thread before commenting on one post tbh
-
@mcc oh that's awesome! back then linux wasn't yet sloppified too, simpler times
next thing i'd like to try on this is haiku. also, there's a lot more performant (and OoO) risc-v soft cores now@mcc my fantasy is something a bit like the mega65, or other retro fpga things (mister), in the reform next shell, and with quickly swappable cores that can share files, so you can try many different old/new/homebrew computers/os on the same set of (your) data. i think linux is mostly not really fun on simple hardware, so i agree with your desire of adopting sth like redox. personally i'd try to do sth as simple as possible but with memory protection (the one thing 90s OSes weren't great at)
