Board support layer about rosflight_firmware HOT 4 CLOSED

I would be totally on board. I wouldn't be opposed to going to c++, since that would make things a lot easier, but moving to C++ would probably require a huge re-vamp of our stuff. Perhaps let's not do that quite yet.

What if instead we used the board.h header like you said above? That seems the most straight-forward. This also would make it easier to switch to the F3 or F4 chips when we are ready. I'm on board, and it shouldn't be too bad. Let's get the controller working first. I think we are about there. I'll go test it right now, and if it works, I'll do a pull request.

from rosflight_firmware.

It sounds like you guys are really forging ahead with this project! I'm especially interested in seeing how you work with the new F3 and F4 flight-control boards. I tried working F3 support into BreezySTM32 but gave up after realizing I was in over my head.

As for going to C++, I'm not sure whether you mean switching to C++ from straight C, or going to C++ from a completely different language. If it's just C -> C++, I wouldn't worry too much: my C++ firmware in Hackflight is really just C with a thin wrapper of C++ classes. Looking back on it, I could just have easily stuck with the original Baselfight approach: a bunch of C modules with static variables and methods for their "private" components, and extern for the "public". Since I'm not creating more than one instance of anything, C++ provides no real advantage. I had to wrap everything in "extern C {" to get it to work with Breezy STM32.

from rosflight_firmware.

Hi @simondlevy, thanks for the input! We'll keep you posted on the progress with the F3/F4 boards, but that's probably not going to happen until a little later down the line.

We were talking about going the C -> C++ route. It's good to get your insight having tried it out. I think the only reason we would do it is to make the code more readable and a little better organized. We may make the switch at some point, but the C version really isn't too bad so we will probably just push ahead with that for now.

from rosflight_firmware.

Completed by #103

from rosflight_firmware.