Servers by jDrones

Dev Call Feb 22, 2021

**Issues & Pull Requests"




1 Like

Attendee count (max): 21

UTC2300 -

  • Documentation for custom mode map in pymavlink
  • Merged!

UTC2303 -

  • No user base so we’re moving the sprayer ID?
    • 216 to 199…
    • We got bitten by this on the EFI message ID changed and it hurt
  • We’ll merge this one but lets try not to do this into the future
  • There’s no reason for this
  • Ranges for numbers have really gone out the window
  • Why move it?
    • Lots of stuff in ardupilotmega.xml could notionally be moved…
      • Worry about extensions being added to these messages without ardupilot’s OK
        • Or new meaning given to extra parameters
        • Which is NOT a problem if it is in ardupilotmega.xml
    • ATM this has to be in common.xml if it is to appear in QGC
  • Should discuss possibility of using all.xml for ground control stations
  • Tridge likes having a separate xml file where we own the messages in it…
  • Peter tridge Amilcar and James to discuss this PR….

UTC2326 -

  • Emit a FENCE_STATUS straight away rather than waiting for streamrate
  • Merged

UTC2329 -

  • Boolean cleanups

UTC2329 -

  • Setting to zero resets the jump tracking
  • 5-jump-loop and then you land and you say waypoint 1 then the jump tracker retains the value across mission runs
    • Setting to zero resets the jump tracker
  • Copter resets the mission on disarm….
  • Additional check for sending in zero
  • Add a comment to point to the other way of doing things
  • Mergeable otherwise

UTC2338 -

  • Need more information on this timestamp ms wrap thing and how it happened
  • GPS time receive going very very large according to PR author

UTC2343 -

  • Tridge is working on this with the user on this high-wind issue

UTC2346 -

  • Move the check into the reboot function instead
  • Remove the infinite loop
  • Need to worry about cygwin

UTC2357 -

  • Add option to turn off plane angle limits and use just q_angle_max
    • When you’re in q modes
  • For vehicles that are fast and can’t afford large angles in fixed wing flight but need them in q-modes
  • Separate function?
  • Tridge will have to come back to look at this one, its not clear

UTC0008 -

  • MdB doesn’t like it…
  • Tridge doesn’t like his lua scripts faulting
  • No GPS at all causes lua scripts to fail if you ask for status on non-existant GPS
  • In LUA you have to check sensor count before checking status
  • MdB thinks we should be protecting against catching invalid data
  • AP_GPS should change to have a boolean on each getter function?
    • Groundspeed() would require a minimum status to be passed in?
      • Some parts of the code are happy with 2D fix some are happy with 3D…
  • If LUA scripts are looking at sensors they’re prone to exit…
    • If battery[1] healthy() then check voltage may exit on startup if your battery monitor hasn’t come up yet (the healthy() call may fault)
    • Set_current_command can fault your script if you ask to set something beyond your number of commands
      • Your script could exit due to a race condition if someone else updates the mission
      • Pcall - protected call
        • Wrap entire script in a pcall….
      • You could pcall a lot to make sure your script doesn’t fault
      • A lot of our examples are unsafe as written
    • Exitting script makes it obvious you’ve made a mistake
      • But exiting in flight is not acceptable…
    • We both return bool and throw exceptions from set_current_waypoint
  • The pcall stuff helps in that you can restart
  • Script failed because a GPS didn’t turn up and it was expected to be there
    • Status was being checked on an instance that didn’t exist
  • Descriptions file is wrong as it is depending on the data
  • Pcalling everything would make for awkward code?

UTC0025 -

  • Merged
  • Send new waypoint on all links

UTC0027 -

  • Useful for debugging
  • Allows RC channel aux options to be triggered in missions and via mavlink
  • Parameter conversion could be a problem?
    • Parachute on 9 could be moved
  • There are checks that make sure you have an interlock and whatnot
  • Whitelist of blacklist?
    • Maybe later?
  • Very likely to be merged - need to make a function boolean

UTC0042 -

  • Move video tx to own directory
    • Which already existed…
  • Merged!

UTC0043 -

  • Running add vtx power to OSD
  • Running out of parameter IDS in current OSD stuff…
  • Can be merged soonish
  • Squilter to poke people on Discord for a merge

UTC0049 -

  • Preserves behaviour in face of incoming changes from Paul
  • This might have unintended consequences
  • Randy is going to have a closer look

UTC0059 -

  • Only send proximity messages where we’ve previously sent them
  • Merged

UTC0103 -

  • Leonard is happy with the squelching of thrust-loss with the parameter
  • Yaw imbalance check
  • Couple more queries but likely to be merged

UTC0106 -

  • I2c bus in scripting
  • Needs MdB to look at it
  • Cap on memory added
  • 4 devices per script per boot
    • Ownptr issue has gone away because of that
  • MdB doesn’t like the example
  • MdB to approve
  • We need something like this if we’re bouncing PRs like Jaime’s

UTC0112 -

  • Notify motor malfunction….
  • Should probably have a mask of failed motors
  • Need extensions to SYS_STATUS for more bits….
  • We really should be using ACTUATOR
    • Flow of how changes are going in is not good…
  • The boolean just tells you you have a problem somewhere on your system
  • A specific hardware failure message?
    • Something like a pymavlink script could be checking that message and looking at what’s broken and acting accordingly…
    • Spreading the data out too much?
      • Consider this is the “high-level” report
      • Another message for details?
  • Use heartbeat’s state?
  • Have a “hardware failsafe”?
  • Motor thrust loss was supposed to be a first step

UTC0126 -

  • Common logging function between for the GSF
  • Different bias estimates in the EKFs can explain the different GSF values

UTC0136 - Plane update

  • 4.0.9 tagged
    • Announcement coming

UTC0137 - Copter update

  • 4.0.7 tagged
    • Announcement coming
  • Bill and amilcar looking at this release…
  • Making progress towards 4.1
    • SCurves into good shape now
      • Some testing done on the weekend
        • Found some issues that have been resolved
        • SplineCurve bug
      • Only known issue is object avoidance
        • Thanks to Rishabh
        • Vehicle is a little more jerky

UTC0139 - no Rover update

UTC0139 - close

1 Like
Servers by jDrones