raddbg UI running in some frankenstein state in X11 and opengl after a weekend of prompts with gemini 2.5 pro exp
TLDR
raddbg UI running in some frankenstein state in X11 and opengl. This was done in 1 weekend with gemini 2.5 pro and me, Danilo, who's never seen the innards of futexes, freetype, X11, opengl or mdesk before Saturday 2pm (start time). I've learned a lot! I'm not suggesting you merge this---
all I wanna know is: what do you think? how horrible is the result really? would you have finished the whole port in a weekend? Roast it!
known issues
- freetype code is derp
- mouse click events? probably a lot more os event issues, least tested
- blur has a weird render artifact
why do this
Ever since Mr Blow mentioned discarding all of linux userspace, the idea sorta just stuck, I think all services should be relegated to net API requests and UI should be Mr Fleury style!
I think it's possible to coax the latest reasoning models into writing UI code raddbg-style, in real-time (well, speed of thought I suppose). They could do so autonomously, and once you have that, well, I think build.sh best describes my aspirations there! :mage_man: plenty of things to build!
@ryanfleury ! Amazing work man! I'd love to help get raddbg working on linux, with remote (server) process debugging capabilities, I hope this is a step in the right direction HAHA!
I don't think we will be able to accept PR or even look at any code that is written with help of LLM. The copyright situation is too unclear with that.
I don't think we will be able to accept PR or even look at any code that is written with help of LLM. The copyright situation is too unclear with that.
I've seen Mr Muratori post about the questionability of AI Labs using a bunch of copyrighted material without any kind of compensation to its authors, what are your thoughts?
This has now all been manually done and is in the codebase now.