Dev Call Dec 12, 2022

Issues & Pull Requests


Copter & Rover

Attendee count (max): 21

UTC2300 - Linux board's don't support all features by LaneaLucy · Pull Request #4783 · ArduPilot/ardupilot_wiki · GitHub

  • Linux boards don’t support all features
    • E.g. d-shot
  • Add documentation to that effect
  • Randy is asked once a day which autopilot to use
    • Thinks generally we should steer people away from them
    • Harder to get started with
  • What’s being sold?
    • Hats for RPi
    • NavIO2
    • Navigator
    • Disco and Bebop
  • We do support them / provide the firmware
  • We could do the extract-features thing

UTC2314 - ExtLibs: ArduPilot: Mavlink: try and pull AP firmware type from version string by IamPete1 · Pull Request #2972 · ArduPilot/MissionPlanner · GitHub

  • Pull AP firmware version from version string
  • The MAV_TYPE isn’t sufficient for MP to know what’s going on
  • E.g. Mode numbers are wrong in dropdowns
  • E.g. param descriptions can be wrong
  • See if we can wedge information into AUTOPILOT_VERSION
    • Flight_sw_version?
      • No space

UTC2326 - Copter: add weathervaneing by IamPete1 · Pull Request #21609 · ArduPilot/ardupilot · GitHub

  • Copter weathervaning
    • Bill wants this tested on Heli
    • Can be merged after that testing

UTC2332 - Plane: if RTL with VTOL active and within QRTL radius and alt then QRTL by IamPete1 · Pull Request #21883 · ArduPilot/ardupilot · GitHub

  • RTL / VTOL / QRTL fixes
  • Merged!

UTC2333 - Copter: Reserve G2 63 for Ship Operations by lthall · Pull Request #22251 · ArduPilot/ardupilot · GitHub

  • Reserve a parameter space for ship operations
  • merged!

UTC2334 - AP_Airspeed:add warning for large offset change by Hwurzburg · Pull Request #22256 · ArduPilot/ardupilot · GitHub

  • Add warning for large offset change
  • Henry tried adding a prearm but it wouldn’t have given a message, so that change was reverted
  • Rebased using interface
    • “Resolve conflicts” is terrible, don’t use it
    • Double-rebase can get rid of the merge commit if you get one accidentally?!
    • “Retry” forgets the “no merge commit” selection

UTC2340 - Payload Place by lthall · Pull Request #22258 · ArduPilot/ardupilot · GitHub

  • Payload place
  • Randy will look at it so Leonard can test it on expensive aircraft

UTC2342 - AP_Airspeed: move to per-instance params by IamPete1 · Pull Request #22297 · ArduPilot/ardupilot · GitHub

  • Per-instance params
  • Overall airspeed enable
  • Renames ARSPD → ARSPD1
  • Changes default?!
  • If first or second type is non-zero in storage it enables the driver
    • This won’t work for 4525
  • There’s a set-defaults call
  • Can we lose the “1” change?
    • Yes
    • Non-numbered ones are a bit confusing
  • Largest size impact is on Rover?
  • Follow-up is support for a third airspeed sensor
    • New airspeed type of “median”
      • Have to be careful a synthetic doesn’t collide with the real
  • Periph costs a lot because of conversion
  • Most parameters aren’t needed on AP_PERIPH
  • So we could more-aggressively remove them
  • Would be nice to do auto-detection but it’s basically not possible due to tube-length-range issues and looks-like-a-baro issues

UTC2355 - Move controllers and PID objects to a real time update system by lthall · Pull Request #22298 · ArduPilot/ardupilot · GitHub

  • PID objects to a real-time-update system
  • First need to get his into master
  • Randy’s flown this
    • Backport to 4.3 with just the fast-task fix
      • Twitch went away
        • Ramp up scheduler loop rate…
        • Psc was still not happy, as expected as this wasn’t the full fix
    • Full fix got rid of twitch and PSC was fully happy
      • Compare position and velocity
        • Just do the multiplication yourself and you can see the slowdown pre-fix
  • Tridge thinks this should go into master now
  • Motors change in 6dof was more extensive than expected
  • MergeOnCIPass, but we need to discuss what goes into 4.3
  • New beta just went out
  • Tridge isn’t inclined to bring this into 4.3.2rc2
    • The smaller patch seems to do the job
    • Maybe in a 4.3.3?
    • If vehicles were running slow then this will affect tuning
    • Would be nice to have some comparative logs
      • Same vehicle, 4.2 back-to-back with 4.3
        • Can’t see why it should be slower
          • Should be faster in a lot of situations
          • If you have an example, please bring it up!
            • Much less likely to bring it back if we can’t see something that needs fixing
    • Low number of testers is problematic
      • Only 30 downloads of Copter beta vs over 2000 stable
    • A backport branch in case people have issues?
    • How long ‘til 4.4?
      • January…
    • Are we still putting fires out in 4.3?
      • Not as good as 4.2 in some ways?
      • Not getting rid of the bugs fast enough?
      • Regressions?
      • Randy’s torn between fixing bugs and introducing features
      • Move back to a slower release cycle?
        • Tridge likes the shorter release cycle
        • Support for new hardware?
        • “Update backport release?”
      • Tridge thinks the shorter release cycles are great
        • The smaller changeset in 4.3 was a great relief
        • Particularly with lack of beta-testing?
        • Beta periods too short?
          • Nobody testing Plane except tridge?
      • Early days of open-source sees much more testing
        • As things get more stable fewer test the bleeding edge and more people using stables
      • [11:21 AM]timtuxworth: I’ve noticed that the testing cycle for a release seems very short.
      • [11:22 AM]timtuxworth: It always seems to me like from hearing there is a beta, I never even have time to run the beta before its already moved on to the next beta or the release
      • [11:22 AM]Peter Hall: We might get more beta testers if the beta cycles are more spread out.
      • [11:23 AM]timtuxworth: I don’t know about before
      • [11:24 AM]timtuxworth: I flew the current Plane beta and crashed (due to pilot error) - sorry no feedback. (edited)
      • [11:26 AM]rmackay9: sadly I need to drop my wife off at the train station shortly…
      • Drop beta releases entirely?!
        • A drop every 6 months
        • Kind of radical!

UTC0036 - add support for mavlink in-progress message by peterbarker · Pull Request #22317 · ArduPilot/ardupilot · GitHub

  • Return in-progress mavlink messages for long-running tasks
  • NeedsTesting label
    • Will test this weekend
  • What happens if you don’t have an airspeed sensor?

UTC0038 - Plane: fix loiter-to-alt for short-distance waypoints by peterbarker · Pull Request #22353 · ArduPilot/ardupilot · GitHub

  • Fix consecutive loiter-to-alt commands
  • Merged!

UTC0041 - Plane update

UTC0044 - Copter/Rover-4.3 issues list · Issue #21711 · ArduPilot/ardupilot · GitHub

  • No Randy…

UTC0046 - Copter: Simulink Model and init scripts for Copter-4.3 by fbredeme · Pull Request #22354 · ArduPilot/ardupilot · GitHub

  • Backport simulink
  • Need Randy

UTC0046 - CAN: added CAN statistics logging for DroneCAN by tridge · Pull Request #22359 · ArduPilot/ardupilot · GitHub

  • Add logging of CAN buses
  • Find bus failures
  • Later we might do separate protocols
  • Broadcast calls aren’t checked
  • MergeOnCIPass

UTC0052 - Copter/Rover-4.3 issues list · Issue #21711 · ArduPilot/ardupilot · GitHub

  • Copter update
  • Twitch fix is in new beta
  • Sporty takeoff
    • Users who think the takeoff is much more aggressive than it used to be
    • non-GPS nav overshoot was a bit scary which needs looking at
    • Poor hover-throttle might explain this
  • Non-gps navigation arming issues
    • Ahrs and gps positions differ
  • Message rate issue
    • Possibly just streamrate parameters?

UTC0056 - Compile AP_Generator_Richenpower in by default by Tabocim · Pull Request #22369 · ArduPilot/ardupilot · GitHub

  • Compile support for the Richenpower in by default
  • “Premium autopilots”
    • Enable it on boards that users are likely to use?
    • FRAM gives 256kB more flash which means to be premium you need FRAM?
  • Make it a LUA script?
  • Won’t be putting this onto boards by default
    • Custom build server and/or OEM compilation setup

UTC0105 - AP_RangeFinder: skip GPIO arming check on analog backend by IamPete1 · Pull Request #22375 · ArduPilot/ardupilot · GitHub

  • Skip GPIO arming check on analogue backend
  • Merged!

UTC0106 - Plane: Auto Landing Refinement by rbeall · Pull Request #22378 · ArduPilot/ardupilot · GitHub

  • Auto-landing refinement
  • Headwind compensation factor currently hard-coded, this adds a parameter
  • Generally looks good, a couple of minor changes

UTC0111 - hwdef: added MatekL431-BattMon support by tridge · Pull Request #22382 · ArduPilot/ardupilot · GitHub

  • Add hwdef for MatekL431-BattMon
  • Related:
  • Why are we doing Ah rather than Wh?
    • Leonard and Paul disagree on which is more useful.
    • Calculate it based on sag-corrected voltage
    • Report both?
      • Put it into the flight controller rather than the nodes
  • Josh will keep looking at things
  • Node-capabilities for deprecating stuff

UTC0129 - Applets:add SportPlaneAirshow schedule by Hwurzburg · Pull Request #22384 · ArduPilot/ardupilot · GitHub

  • Aerobatics Schedule update
  • Merged!

UTC0130 - DroneCAN: update submodules by tridge · Pull Request #22387 · ArduPilot/ardupilot · GitHub

  • Update dronecan submodules
  • Merged!

UTC0131 - close