Maxime Pinard

I think it's a good idea, this project structure is also followed by SFML, it will be more consistent in the includes. But we can't have full mandatory consistency, ImGui...

The problem putting `imgui.h` in `/usr/include/` is that someone installing ImGui or another binding wight override it with a future/past unsupported version. Putting `imgui.h` in `/usr/include/imgui/` can lead to the...

I think it is worth it to make a breaking change to be more consistent with the other back-ends.

Yes, I wanted to do some benchmarking to compare with at least `std::vector` and `boost::dynamic_bitset`, maybe using Catch2 benchmarking functionality as it is already a dependancy for the tests. I...

I'm back but now I have a summer job so I have little free time left to work on this... I work on it when I can but it will...

I decided to use [google/benchmark](https://github.com/google/benchmark) because I may need the csv/json format output later, to generate performance comparison graphs (maybe with some cool LaTeX magic). I created a new repository...

Yes, it will add an undefined behaviour (assert in debug) for bitsets longer than an ``unsigned long long`` but it is an useful function, I will add it when I...

I agree, any hope it will be fixed?

I made a Mandelbrot set visualization with mouse-centered zoom/de-zoom:  I also took the liberty to change from `mouse_data.y = (io.MousePos.y - p1.y) / size.y;` to `mouse_data.y = (io.MousePos.y -...

Following the Mandelbrot set visualization, this time I made a Julia sets visualization:  Inspired by [this Wikipedia gif](https://en.wikipedia.org/wiki/File:JSr07885.gif), the Julia sets represented are generated by ``` z^{2} + 0.7885...