Michele Caini

@alexshpilkin I've no way to test a fix to be honest. If you know already what to change and can also verify it, can I ask you to open a...

Ping. The issue is starving. I cannot reproduce the problem, I need help in case and I don't have any way to test it eventually.

I don't but it might be an option actually. Should be easy to adapt the dense map. Do you want to give it a try?

Updated the base branch to `wip`. It fails to compile on all platforms though (yeah, the toolset v141 is always a pleasure to work with in 2022 😅).

The one in `view.cpp` is due to an error of mine that is already fixed on `wip`, so don't worry. As for the other one, it's in a new file,...

More or less. In fact, you can _compact_ the storage. However, as long as the type is at least copyable, it works as well. The challenge is with non-copyable and...

Wait a moment. Non-movable types [are already supported out of the box](https://godbolt.org/z/vTab6sn7W). In fact, `is_move_constructible_v` and `is_move_assignable_v` are true for non-movable types that are copyable. Does the problem arise when...

I see, yeah. So, the problem isn't about non-movable objects (this is easy solved and I'll do). It's about non-movable and non-copyable types. Slightly more complex but still feasible with...

> Added a bit about optionally allowing view to return disabled entities. I think this goes without saying because of the implementation I've in mind. When you iterate the storage...

Hi, first of all: thank you very much!! Really interesting read and overview. I see the point of the change too. I don't think we can merge it as-is because...