"Vulkan Next" Is In Active Development

[Azpegath]

What can you do, Khronos doesn't do tools. Microsoft, on the other hand, is pretty good at this (in part because everything is done in house and they don't have to coordinate with other parties).

I think this is meant more for professional usage, where the percentage will be "slightly" higher than 1%. Fwiw, I think desktop usage of SLI/Crossfire is well under 1%. But it doesn't really matter.

Well, if more of you guys had actually watched the stream (my apologies if you did) you would have seen that Tom raves about the amazing community that has sprung up, and is SO glad that they've open-sourced everything (thanks to the constant nagging from Christophe Riccie, a prominent developer and FOSS-proponent at Unity and GLM fame, among other things). He also talks about Khronos complete lack of funds, but that they are working on being able to support the development of much improved tutorials and tools. During this, they keep thanking Baldur for doing RenderDoc, which I've only heard great things about.

[humbug]

Its like comparing DX12 against OpenGL. Its sad thats not much faster as D3D11. Nothing to boast.

their vulkan renderer is a work in progress and under active development. You only get it if you opt into public beta. The openGL and directx renderers are mature. So the performance is not sad at all...

[plonoma]

generalized render-pass/sub-pass dependencies

If GPU, graphics developers could freely specify passess and sub-passess in a render/compute loop. And by specifying the order and dependencies in an elegant and straightforward way, this would be great!
I would love to see freeform pipelines able to use vertex, fragment, tesselation, compute, other passess in any arbitrary order. Really allows the game engine programmers to code the pipeline how they see fit.
(Specifying the order by using the order of the passess in the code should definitely be in it.)

But surely using compute passess and sub-passess in the render loop would be part of this generalized ordering mechanism as well right?
And what about asynchronous shaders? Will they be specified?
Will it be possible to specify using a result from an asynchronous shader when the result is ready and use either a placeholder result or alternative execution path in the shader when the result is not ready yet?

[pal666]

"a rigorous memory model"

Uh, wasn't the whole point of these shiny new APIs the fact that they wanted to get rid of this?

you seem to be confused with what memory model is
here is education for you https://en.wikipedia.org/wiki/Memory..._(programming)

[pal666]

What about better debugging tools, better example codes, mature tools, etc?

i assume that stuff do not require api changes

[pal666]

I know having support in Visual Studio would go a long way for many devs.

will those devs develop for linux?

[pal666]

Yeah I think they need to be convincing devs to use Vulkan

what devs? all major engines are part of khronos vulkan developers