achal

> Therefore IGPUCommandBuffer::emplace and possibly even the deleteCommandSegmentList (as a function taking a head segment parameter) should move to IGPUCommandPool. For these: ```C++ m_cmdpool->m_commandSegmentPool.allocate(IGPUCommandPool::COMMAND_SEGMENT_SIZE, alignof(IGPUCommandPool::CommandSegment)); ``` and ```C++ m_cmdpool->m_commandSegmentPool.deallocate(currSegment, IGPUCommandPool::COMMAND_SEGMENT_SIZE);...

## OpenGL Command Buffer Baking ### What is happening now - `SOpenGLContextLocalCache` is cached in the command buffer's `COpenGL_Queue`'s thread internal state (`ThreadInternalStateType`). - The cached `SOpenGLContextLocalCache` is modified (_eventually_...

> Then for every command that doesn't do weird variable length shenanigans, like CDrawIndexedIndirectCountCmd, you could just do Wouldn't every command be just fixed size? For example, even for `CPipelineBarrierCmd`...

## Descriptor Lifetime Tracking GPU resources, referenced by descriptor sets, go out of scope and get destroyed before we get the chance to bind the descriptor set and use them....

Merge to `master` TODOs: - [x] Merge [achalpandeyy/blur](https://github.com/achalpandeyy/Nabla/tree/blur) with the current `master` and make ex25 compile and work. - [x] Get ex25 running on old API. - [x] Merge with...

1. Is it plausible to make `ISwapchain` templated on the type of pointers (raw or refctd) it uses to keep its images? If the images are created externally (Vulkan's Presentation...

Done here https://github.com/Devsh-Graphics-Programming/Nabla/pull/345/commits/06f524ff0a0ac757e7a190735f47aac5dc940d83

> but obviously a bad example because engine max descriptor set count is hardcoded to 4. Okay I will change this to a flat array. > > Commands affected by...

> > > but obviously a bad example because engine max descriptor set count is hardcoded to 4. > > > > > > Okay I will change this to...

Hi! Is anybody looking into this? I'd like to take a stab at it.