Dev call May 27 2024

Issues & Pull Requests

Copter & Rover


Attendee count (max): 17

UTC1100 - Add multiplier value for GPS_STATUS.satellite_azimuth where 255 means 360deg by shancock884 · Pull Request #885 · ArduPilot/pymavlink · GitHub

  • Add a multiplier for satellite azimuth
  • merged!

UTC1103 - common.xml: add CAPA indicating mission-slot-0 is home by peterbarker · Pull Request #360 · ArduPilot/mavlink · GitHub

  • Could do a special NAV_HOME mission item type too
  • Parsing tlogs - don’t have capability bits so you don’t know what home is
  • GCSs which can’t read capabilities can’t work (e.g. read-only GCSs)
  • Still don’t know whether to add home at slot 0
  • Probably need the capability flags
  • Introduce NAV_HOME at the same time
    • Can NAV_HOME be anywhere than slot-0?
    • Nav_home an be displayed specially on the map
  • Will come back to it next week

UTC1119 - Copter: Adding a class name by muramura · Pull Request #24621 · ArduPilot/ardupilot · GitHub

  • Add namespacing to millis/micros call
  • No, it makes things smaller+faster

UTC1121 - AP_HAL_SITL: remove deprecated --uartX options by tpwrules · Pull Request #26156 · ArduPilot/ardupilot · GitHub

  • Remove deprecated --uartX options
  • Merged!

UTC1123 - Copter: Autotune performance and safety enhancements by lthall · Pull Request #26245 · ArduPilot/ardupilot · GitHub

  • Copter performance and safety enhancements
  • Randy’s tested it and asked for it to be more lenient in terms of “can’t level” errors
  • Trying to reduce the occurrence of bad autotune outcomes
    • By making it tighter in a number of ways
  • Andy’s tested it
  • Leonard’s happy with it
    • Wants it merged so he can get even more feedback on it
  • Can’t restart an autotune?
    • Tuning on a windy day
    • It refused to tune
  • Reposition?
    • If the user leaves the aircraft moving at 15m/s after you’ve repositioned then
  • Tested under loiter?
    • Maybe not?
  • gcs().send_text should be GCS_SEND_TEXT
  • Logs that tridge sent through?
    • Tridge is re-testing
    • Failing in RealFlight now, which is good
  • Will need to tune in lower wind with this new code
    • And that’s a good thing
    • These are supposed to be tuning flights and should be done in ideal conditions, ideally
  • Passes Heli autotest which is good
  • Need to adjust the wiki to really express what “failed to level” actually means

UTC1144 - ChibiOS: UART: Add support for RS-485 Driver Enable RTS flow control by IamPete1 · Pull Request #26930 · ArduPilot/ardupilot · GitHub

  • Add support for RS-485 flow control
  • Has been tried on vehicle and it works well for PH
  • Can simplify the torqeedo driver to remove the pin
  • Description should be in metadata
    • “Nobody reads the Wiki” – Henry

UTC1150 - Start Surface Distance Library by MattKear · Pull Request #26933 · ArduPilot/ardupilot · GitHub

  • Surface distance library
  • Approved!
  • Merged!

UTC1155 - AP_HAL_ChibiOS: add hwdef for Flywoo/Goku F405 HD 1-2S ELRS AIO v2 by rotorman · Pull Request #27036 · ArduPilot/ardupilot · GitHub

  • Adds another Flywoo/Goku flight controller
  • Merged!

UTC1157 - Create common GSOF library by Ryanf55 · Pull Request #27058 · ArduPilot/ardupilot · GitHub

  • Create common GSOF library
  • Why just “GSOF”?
    • Maybe “TrimbleGSOF” instead?
  • Just moving code
  • Run Valgrind, ping on code review

UTC0003 -

  • Give hint about installing stuff rather than Python exception when running github-release-downloader
  • Merged!
  • Need a patch to check only a single file at a time

UTC0004 - Copter: ModeGuided: add an option_is_enabled method, use it by peterbarker · Pull Request #27141 · ArduPilot/ardupilot · GitHub

  • Add an option_is_enabled
  • Can be merged once comment and bug fixed

UTC0007 - AP_Scripting: remove support for REPL by IamPete1 · Pull Request #27146 · ArduPilot/ardupilot · GitHub

  • Remove LUA REPL support
  • Merged!

UTC0010 - GCS_MAVLink: default support for BATTERY2 sending to off by peterbarker · Pull Request #27148 · ArduPilot/ardupilot · GitHub

  • Stop compiling in support for sending BATTERY2 message
  • merged!

UTC0012 - Tools: Install uxrceddsgen in Ubuntu by Ryanf55 · Pull Request #27150 · ArduPilot/ardupilot · GitHub

  • Install uxreddsgen in Ubuntu
    • In prereqs
  • Peter is suggesting we put the artifact into the ArduPilot tree
  • “People don’t read the Wiki” –Ryan
  • We’ll merge as-is and see what happens
    • Check jre behaviour first

UTC0022 - waf: use to get boards list by peterbarker · Pull Request #27153 · ArduPilot/ardupilot · GitHub

  • Move rough parsing of hwdef files into chibios_hwdef
  • Merged!

UTC0026 - AP_DroneCAN: Serial: map baudrates so param works as expected by IamPete1 · Pull Request #27160 · ArduPilot/ardupilot · GitHub

  • Map baudrates in AP_Peiph so param works as expected
  • Merged!

UTC0026 - AP_DDS: Use common quaternion initialization function by Ryanf55 · Pull Request #27161 · ArduPilot/ardupilot · GitHub

  • Use common quaternion init function
  • Don’t want to inline the initialise function
    • fixed
  • MergeOnCIPass now

UTC0029 - AP_Scripting: added notch_switch example by tridge · Pull Request #27163 · ArduPilot/ardupilot · GitHub

  • Added notch_switch example
  • Might be a starting point for a change-my-tune LUA script
  • MergeOnCIPass

UTC0036 - Filter: added test of phase lag vs attenuation by tridge · Pull Request #27164 · ArduPilot/ardupilot · GitHub

  • Add test of phase lag vs attenuation
  • Shows correlation between phase lag and attenuation
  • Why the stair-step effect?
  • Leonard thinks this isn’t actually correlation
  • The range is wrong, should be 40dB to the right
  • People often put in 40db in the filter tool
    • Taking the noise at that point well below the noise floor at that point
    • You’re adding phase lag for no good reason
    • Over-filtering it!
    • Don’t bring it to 30dB below ambient, reduce to 10dB below ambient
    • End up at 15dB
    • Reducing phase lag
    • Works better if you have lots of notches
      • particularly with multi-notch
  • Different sources are sinusoids at different frequencies
  • Large number of overlapping notches?
    • Argument to be made for lower gains
      • Particularly on an octo
      • Maybe 10dB
      • Could halve the phase lag and might be enough given the number of overlapping notches
  • Stepping might be due to the crossing problem
  • Tridge would like a tool showing how much margin people have
  • Leonard and tridge will chat

UTC0106 - Moved from using new to new(std::nothrow) by tridge · Pull Request #27169 · ArduPilot/ardupilot · GitHub

  • Move to new(std::nothrow)
  • Clang++ swallows the -fcheck-new compiler flag and ignores it
    • So compilers are clever and omit the nullptr checking on new
    • So a flag is added to say “don’t do that”
    • So the clang++ maintainers take that flag and ignore it
  • It is unsafe to fly with a clang++ ArduPilot binary ATM
  • Lifted the no-exceptions thing up
    • Can’t easily use it on Linux as libc uses exceptions in places
  • Costs flash
    • Probably the new argument
  • Rejig to get zero flash cost on ChibiOS
  • Randy’s looking forward to having the QURT port for the tightly integrated ModalAI device
    • Baro is up, IMU is up
    • Mavlink next….

UTC0131 - SITL: update to work on newer cygwin versions by tridge · Pull Request #27170 · ArduPilot/ardupilot · GitHub

  • Allow newer cygwin to work
  • Merged!

UTC0134 - AP_Scripting: added GPS yaw binding by tridge · Pull Request #27178 · ArduPilot/ardupilot · GitHub

  • Add GPS yaw binding
  • Needs to pass CI

UTC0137 - Copter/Rover-4.5.0 issues list · Issue #26103 · ArduPilot/ardupilot · GitHub

  • Copter update
  • Internal error from Tim
    • Have a good log for that
    • Leonard will look at
  • Not a lot else new
  • Release notes for Rover talked about ICE enhancements
    • But not supported for Rover…
  • QGC testing is ongoing
    • Henry’s uncovering lots of issues
    • Tonnes of HereLink issues
      • Contact Randy if you come across some, he wants to tighten things up there
  • Randy will be away for a week-and-a-half or more

UTC0140 - Plane update

  • Noticed IOMCU reset on bdshot on an IOMCU
    • Something in Andy’s area
    • Seems to be harmless

UTC0145 - close