Darío
Darío
I'd like to add them, but the main problem is that the RT64 renderer is written in D3D12 only at the moment. This can be solved as soon as a...
Also, AMD is already supported in Windows! It's just the denoiser won't work (which will probably be solved by using a denoiser that doesn't use a CUDA-only library).
That's alright, I'll leave it open for anyone curious to check out the reasons why Linux is not supported at the moment.
The current plan is to just have DX12/Vulkan backends in the future RT64 version that will end up replacing the PC renderer on sm64rt (as well as be a standalone...
The RT64 rewrite already has Vulkan support so I'll likely just port sm64rt to that later. https://twitter.com/dariosamo/status/1716189711889322005
Not yet, still got a lot of work to do and rewriting sm64rt to use it will take even longer as well. Alternatively, if we don't want to rewrite a...
Pretty please.
> _Edit: On second thought, this will not be trivial. It seems to rely on Vulkan-specific properties that we don't expose through the RD right now. So implementing @darksylinc's solution...
> This is a vulkan-specific workaround though. Shouldn't the flag be set at the Vk implementation level? It's a bit of a mixed bag as the workaround _must_ be implemented...
I've pushed a new version of the workaround detection based on driver version instead. For this I moved the logic specifically to be a part of the Vulkan driver. The...