PX4 Dev Call: March 04, 2020

rroche · February 26, 2020, 6:09pm

March 04, 2020, 4pm GMT (8am GMT-8, 5pm GMT+1)

Meeting ID: 946 175 205
Join using your mobile/desktop

Call in using your phone, find your local number: https://zoom.us/u/aetSYKbiMF

Quick statement about feedback on the dev-call
Status update by component
Development focus
Open section for the community (Please add your agenda items)
- Anyone can propose topics for discussion

Backports from @dagar and @david_s5

@david_s5 Planning to improve DMA for SPI.

TODO: Figure out testing for regressions

TODO: Find reviewer for this PR

There will be a release branch and beta on March 9. for testing.
How to call for community testing on the released beta?
Proposed: Forum post calls for comments about issues AND successful flight testing reports including logs because it’s a lower entry barrier. Include a link to documentation on how to work with GitHub issues.
We need to go over release notes draft

Beta Testing the Release

Beta announcement (create a thread on the forums)
Clear instructions on how to test features and how to report back to the team
Project board https://github.com/PX4/Firmware/milestone/32
(considering organization-wide kanban: https://github.com/orgs/PX4/projects/16)

Release Timeline

Acceleration feed-forward: offboard issue, what do we support?
- No clear documentation on what’s supported. ROS CI should work, MAVSDK tests should work, additional testing is nice to have. Discussing with @JulianOes & @Jaeyoung-Lim because they had offboard problems with and without the pr as well.

Attitude setpoint generation: what do people expect from yaw/heading in mission and manual position flight?
No clear answer which means the pr will be continued with the assumption heading is the angle from north to the vehicle body x-axis when roll and pitch are completely level.

How do we architecturally handle Libraries which are part of a module on flash constrained targets? #ifdefs and dummy interfaces?
https://github.com/PX4/Firmware/pull/14279
There’s no simple design pattern. David had some suggestions which are used by NuttX but it’s not clear if they’re more maintainable than the ifdef dummy. Julian K. had a further idea to have the function calls wrapped with a compile time decision if they are forwarded but it might not get stripped out because of the library still being present as a member. Will be tested and we also ask Daniel next week.

Toolchain setup improvements
Proposition to have minimal output and progress bar for setup script: high effort for the benefit. https://github.com/PX4/Firmware/issues/14249
@JulianOes is collecting python issues and creating a troubleshooting page for typical python environment issues.
Would be nice to consolidate the common parts of the ubuntu and arch script.
What’s definitely missing is using the scripts in the docker container creation instead of replicating the setup there.

@dagar (On vacations) What support is needed?

UAVCAN GPS support
CUAV Nora board
How do we architecturally handle Libraries. follow-up from this week by @jkflying

Need help from the community finding hardware to test PR14269 Navio2, BBB, OCPOC
TODO: Found a problem with k66 where it won’t arm @david_s5 is looking into this

Let me know below if I failed to capture anything the right way, and if there are any updates, or you have feedback on the call format.