Jump to content

  • Log In with Google      Sign In   
  • Create Account


Member Since 08 Sep 2011
Offline Last Active Yesterday, 10:39 AM

#5250664 Does anyone know which OpenGL state did I screw up?

Posted by Mona2000 on 05 September 2015 - 01:28 AM

Fill mode. Second argument in glPolygonMode, should be GL_FILL instead of GL_LINE.

#5248752 sRGB on diffuse textures or gbuffer color texture?

Posted by Mona2000 on 25 August 2015 - 07:22 AM

Also that whole part about D3D not having integer formats is complete hogwash.

#5248419 How to work with legacy textures using DirectX11?

Posted by Mona2000 on 23 August 2015 - 03:08 PM

8bppIndexed: I guess it's not supported and needs to be manually decoded, but I'm not entirely sure.

16bppArgb1555: DXGI_FORMAT_B5G5R5A1_UNORM. Requires feature level 11.1 and Windows 8, else it needs to be converted manually.

16bppRgb555: I only count 15bpp. If that's the case same as above (alpha is ignored), if you meant 565 then DXGI_FORMAT_B5G6R5_UNORM. Same requirement as above.

24bppRgb: not supported, needs to manually converted to 32bpp and loaded as DXGI_FORMAT_B8G8R8X8_UNORM

#5247370 Vulkan is Next-Gen OpenGL

Posted by Mona2000 on 18 August 2015 - 06:21 AM

Interesting links. Basically the only big change between Mantle and Vulkan is the render pass grouping, for tiled architecture friendliness.

#5247079 Vulkan is Next-Gen OpenGL

Posted by Mona2000 on 17 August 2015 - 05:08 AM

@mona2000 you're still coding DX only? Sounds like a endangered species to me. Good luck with that...


PS: Linux is also PC


Nice assumption, also incorrect; I work with both D3D and OpenGL. I know it's mind-blowing to think that somebody who works with OpenGL doesn't worship it blindly, but I can assure you it sometimes happens.
I'm interested in Vulkan for its potential to unify several platforms, and I don't like their current strategy because it looks like it's gonna lead to an even more fragmented ecosystem instead (from Windows: D3D, Linux/Mac/iOS/Android: OpenGL to Windows: D3D, Mac/iOS: Metal, Linux/Android: Vulkan). Oh well.

PS: unfortunately Linux is insignificant and will remain like that until some people don't change their destructive attitudes. Funnily enough, your endangered species threat reminds me of "Linux is the future (of desktops)", a phrase that I've been hearing for the last 25 years or so.

#5246883 Vulkan is Next-Gen OpenGL

Posted by Mona2000 on 16 August 2015 - 06:21 AM

I would say it's the other way around, if they were shortsighted they would just have released long ago. They're holding off to get 10 different vendors to agree on compliance that fits 20 different software teams, in order to not have to redo everything again.

That's exactly what they said of Longs Peak.

#5246877 Vulkan is Next-Gen OpenGL

Posted by Mona2000 on 16 August 2015 - 05:36 AM

It's very clear by now that Vulkan is Mantle 1.1, so I don't get how they managed to get beaten by MS yet again.

Is extreme shortsightedness a requirement for joining Khronos? Do they fight at meetings instead of working?

#5244714 Error: pixel shader must minimally write all four components of SV_Target0

Posted by Mona2000 on 05 August 2015 - 01:30 PM

Well then, how about you tell us the correct way to fix his issue?

I would if I could, but there is no issue in the code he posted; it compiles perfectly fine for me. Perhaps he is compiling it with the wrong target (D3D9on11 or however that thing is called)?

#5244690 Error: pixel shader must minimally write all four components of SV_Target0

Posted by Mona2000 on 05 August 2015 - 11:22 AM

What do you expect to happen with that shader? You have to actually write a pixel out, i.e:


Wrong. If you are rendering to a depth buffer with no render targets bound you still need the pixel shader for alpha testing, but there is no need for any return value.

#5244568 Including headers from Windows 10 SDK

Posted by Mona2000 on 04 August 2015 - 05:12 PM

Thats because since DX11s release D3DX has been deprecated and shouldnt be used, if you want  a replacement for this look at DirectXToolkit at : https://directxtk.codeplex.com/

d3dx12.h is a helper library that has nothing to do with old D3DX and/or DirectXTK.

It's not included in the Windows SDK, you can find a copy here.

#5244049 SharpDX: MRT problem

Posted by Mona2000 on 01 August 2015 - 02:50 PM

Does it give you an error if you use SetTargets instead of SetRenderTargets? If yes, which one?

#5244011 Safety vs Efficiency

Posted by Mona2000 on 01 August 2015 - 10:31 AM

Continuing after an assertion is such a brilliant idea, what could possibly go wrong by running the program in an invalid state..


From personal experience: I worked under a "senior" developer who thought allowing the user to continue after assertions and unhandled exceptions (..) was a good idea. I don't remember one day where I didn't waste my time investigating bogus bug reports that only existed because of invalid state. Seriously, fuck that.

#5243371 Cheap Rendering Tricks Used in Game Industry?

Posted by Mona2000 on 29 July 2015 - 08:27 AM

On a side note, the Wii didn't have hardware support for user clipping planes - so that itself required another trick! We had to make a texture where the top half was black, and the bottom half was white, and then plug this texture into the alpha-testing hardware, and compute it's texture coordinate as "vertex.position.y - floorHeight", which would then either pass or fail the alpha test depending if you were above/below the floor, simulating a clipping plane ohmy.png

Isn't it cheaper to do that with the projection matrix?

#5242767 Issue with Cubemap Shader Resource View Mismatch

Posted by Mona2000 on 26 July 2015 - 09:32 AM

Basically you bind diffuse to slot 0, normal to slot 1, displacement to slot 2, then reflection to slot 1 again.

#5242712 steam hardware survey

Posted by Mona2000 on 25 July 2015 - 07:39 PM

Err, they've been doing that for like 5 years?