fw_att_control: move ecl/attitude_fw into fw_att_control #14284
Conversation
|
@sfuhrer @RomanBapst once this is in let's refactor it in the style of |
There was a problem hiding this comment.
I totally agree with moving it out of ecl, thanks for initializing that @dagar .
I made a quick pass on the code and sitl tested it with a plane and a standard vtol, looks right. I could also give it a quick spin on a real vehicle (VTOL) if you think that's already necessary at this stage.
There is no 1.11 release branch yet right? It that's planned in the (very) near future I guess we should wait with the FW refactoring we start with this, or how do you feel about it?
| @@ -0,0 +1,145 @@ | |||
| /**************************************************************************** | |||
| * | |||
| * Copyright (c) 2013 Estimation and Control Library (ECL). All rights reserved. | |||
There was a problem hiding this comment.
@dagar is it okay to keep the wrong header for now? Or should we at least change them, plus maybe also the file names (remove the "ecl" from them?)
There was a problem hiding this comment.
I think it's fine. They'll also be refactored out of existence at some point.
|
What is keeping this PR from being merged? It concerns me if things just stop for a week. This makes the whole team really slow. We really need to be working on maximum 3-4 things at any given time, but get those things done and merged quickly. |
In this particular case we got blocked by the yaw estimator (EKF-GSF) merging into PX4/ecl master before the required change here (PX4/PX4-ECL#765). |
|
Other ECL change is out of the way, rebasing this PR to get it in. |
- this was never maintained as a separate standalone project - PX4/PX4-Autopilot#14284
dagar
force-pushed
the
pr-fw_attitude_control_ecl
branch
from
65f1a13
to
fab55a2
Compare
Let's merge this back into PX4/Firmware. A large proportion of the code simply goes away if the work is handled in place.
ECL side (required first): PX4/PX4-ECL#765
First step is to copy in place as is, then we can incrementally refactor easily with the submodule gone.