I'd like to ask HN a very vaguely related question. I need to get a self-hosted runner (for GitHub Actions) that is capable of running Windows ARM64. What are my options other than buying a machine and do everything manually? Are there any service providers that offer Windows ARM64 VMs? I can only seem to find options for Linux.
In case you don't already know, there are Github-hosted runners that run Windows arm64 [1]
Also, it's not what you're asking, but self-hosted runners are a security nightmare if you don't have the hardware to completely isolate them from your local network.
Yeah and (1) the codegen produces massive headers that slow compilation of anything that touches them (2) the generated classes are really awkward to use. Not a big fan of the experience of protobuffer generated code in a large C++ code base.
It's lead to a huge layer of adapters and native c++ classes equivalent to the protobuffers classes to try and mitigate these issues.
I wish GPU vendors would stick to a standard terminology, at least for common parts. It's really confusing having to deal with warps vs wavefronts vs simd groups, thread block vs workgroup, streaming multiprocessor vs compute unit vs execution unit, etc...
> Travelling to a 3rd world country with genuinely poor people can help, a bit.
I agree with this. For anyone who's young and can afford, I wholeheartedly recommend travelling to a poor country. It will truly give you a lot perspective on the world and appreciate your life more.
Having said that, I often see people grossly misunderstanding that just because you visited a poor country you understand the lives of the people you saw (e.g. their cultures, opinions and lifestyle). It really does not. You only get a very superficial glimpse, but it takes years to embody their experience in your mind.
> We are collaborating with the Flutter team at Google to bring Impeller, their GPU first renderer, to .NET. That work is already in progress and as it lands, the MAUI backend will inherit those gains.
Interesting, I wonder how good Impeller is and if it's actually better than the new Graphite backend of Skia.
Predictable performance: Impeller compiles all shaders and reflection offline at build time. It builds all pipeline state objects upfront. The engine controls caching and caches explicitly.
or as described here [2]
Flutter’s Impeller renderer outperformed Skia. Impeller eliminates runtime shader compilation stalls, delivering lower frame times and more stable performance. For animation-heavy, graphics-rich apps, enabling Impeller significantly reduces jank and provides a smoother user experience.
I wonder if Mojo can be adapted to be used with graphics. It'd be a dream if we had one single unified API instead of the WebGPU, Vulkan, DirectX and Metal hell.
> Hardware vendors recognized the long-term benefits of a unified software ecosystem, but in the short term, they were fierce competitors. This led to subtle but significant problems: instead of telling the committee about the hardware features you’re working on (giving a competitor a head start), participants would keep innovations secret until after the hardware shipped, and only discuss it after these features became commoditized (using vendor-specific extensions instead).
I wonder whether this same point is what's holding Vulkan back, which always seems to be trailing behind DX12 when it comes to innovation.
reply