A nice thing with std::gfx is that there is surprisingly little needed to build a graphics library now. I got an LLM translation to C3 of @SebAaltonen’s NoGraphicsAPI here: github.com/lerno/nogfx_c3 with std::gfx handling what the API doesn’t do, like keyboard and windows.
A lot of shit we are dealing with in the modern computer systems is an adversarial complexity that serves no purpose except making things harder for the benefit of the system's owners. The main problem is that lot of people believe this complexity is necessary.
I’m looking forward to releasing C3 0.8.5 with the new std::gfx.
It’s not that it’s revolutionary: it isn’t, any language could add this. But it’s taking a stance and having an opinion about what a stdlib should have, and that is somehow satisfying.
It was created by C3 contributor Manu Linares who spent time not just making the compiler run in the browser compiled to WASM, but also making the editor and the graphical parts happen.
With the blog post finally done, C3 0.8.4 is officially out: c3-lang.org/blog/#c3-084-the…
(It's been available for over a week already, but without a blog post it's not quite done)
NoGraphicsAPI is released (MIT license)!
It's a low-level Vulkan 1.4 wrapper using latest extensions (BDA, descriptor heap, device address commands, unified image layouts) to get as close as possibly to my No Graphics API blog post and SIGGRAPH 26 talk.
github.com/sebbbi/NoGraphics…
From the requirements of backend role:
> - Backend (for example Go, Java, Kotlin, Python or similar)
> - Frontend, all technologies
"All technologies"💀 So, all JS frameworks ever created. For a backend role. Such a well written ad.
I don't mind doing Java, but if I see a job description requiring "in-depth experience with Spring Boot and the Spring ecosystem", it's an immediate hard pass.
Let me share my completely uniformed take on DLSS5: it’s inevitable that added rule-based computation eventually will reach the point where there is increasingly little progress and other solutions are searched for. The standard example is trying to create an AI for Go. >
Computation from rules can be incredibly effective, but computing from rules cannot scale to infinity. This is what building AI for Go showed. And I think this is what graphics is now running into: when progress has come so far that neural networks are more effective than rules.>
> Computation is the natural starting point, it is only when computation becomes suitably complex that there is room for alternatives. It exactly because we’ve got so far that we need new a new approach to get further. Inevitable because we want progress.