Let me clarify one thing: The approach we take is a pure userspace solution, it isn't related to NuttX kernel and no plan to upstream it.
That is good to know. Thank you. I will sleep better tonight.
The major difference is where to put L2CAP/HCI code(userspace v.s. kernelspace) and it isn't too hard to migrate to socket interface once the NuttX kernel has the full and stable L2CAP/HCI support. My question is how to finish the Bluetooth stack on the userspace? It will be a huge work to write from scratch(BT, BLE, BLE mesh and BLE Audio).
Since NuttX does not support a user-space Bluetooth stack, I will leave this up to your internal implementation.
I think that Xiaomi is missing a tremendous opportunity. The reason that most corporations get involved with open source projects is to benefit from the open source support from the community. But it always seems the Xiaomi goes their own way and implements things all on their own without discussing with the community. That is Xiaomi's loss. You are losing the power and strength of the open source project.
If you, Chao An, I, Brennan, Matias, and others all worked together with a common goal and under common architectural principles, I think that we could accomplish a lot. Much more than Xiaomi could accomplish by themselves. But that requires, planning, discussion, agreement, and conformance to a common architecture and roadmap.
Yes Bluetooth is a big job. It is a lot for one person. But if we share the project as a true open source project, we can do most anything. There is a lot a strength and talent on this project that you are ignoring.
For example, I offered to implement the Bluetooth protocol sockets this morning (my time). Matias is considering a Gatt server.
