Updated HLSL readme

shaders/hlsl/README.md

@@ -1,18 +1,10 @@
 ## HLSL Shaders
 
-This directory contains a fork of the shaders found in [data/shaders/glsl](https://github.com/SaschaWillems/Vulkan/tree/master/data/shaders/glsl), re-written in HLSL.
-These can be compiled with [DXC](https://github.com/microsoft/DirectXShaderCompiler) using the `compile.py` script.
+This directory contains shaders using the [HLSL shading language](https://docs.vulkan.org/guide/latest/hlsl.html). These can be compiled with [DXC](https://github.com/microsoft/DirectXShaderCompiler) using e.g. the included `compile.py` script.
 
 ### Known issues
 
-- specialization constants can't be used to specify array size.
+- Specialization constants can't be used to specify array size.
 - `gl_PointCoord` not supported. HLSL has no equivalent. We changed the shaders to calulate the PointCoord manually in the shader. (`computenbody`, `computeparticles`, `particlesystem` examples).
-- HLSL doesn't have inverse operation (`deferred`, `hdr`, `instancing`, `skeletalanimation` & `texturecubemap` examples).
-- `modf` causes compilation to fail without errors or warnings. (`modf` not used by any examples, easily confused with fmod)
-- In `specializationconstants` example, shader compilation fails with error:
-    ```
-    --- Error msg: fatal error: failed to optimize SPIR-V: Id 10 is defined more than once
-    ```
-  When multiple constant ids are defined and have different types. We work around this problem by making all constant ids the same type, then use `asfloat`, `asint` or `asuint` to get the original value in the shader.
-- `gl_RayTmaxNV` not supported. (`nv_ray_tracing_*` examples)
+- HLSL doesn't have inverse operation (`deferred`, `hdr`, `instancing`, `skeletalanimation` & `texturecubemap` examples), these should be done on the CPU
 - HLSL interface for sparse residency textures is different from GLSL interface. After translating from HLSL to GLSL the shaders behave slightly different. Most important parts do behave identically though.