Dev Call Feb 03, 2026

Development Team
1

CI Workflows

Release Update

Issues & Pull Requests

2

Attendee count (max): 20

UTC2300 CI workflows ci · Workflow runs · ArduPilot/ardupilot · GitHub

  • Working better with the correction from last week

UTC2311 - release update ArduPilot 4.6 issues list · Issue #28612 · ArduPilot/ardupilot · GitHub

  • SRTM Terrain generation discussion

UTC2320 DroneCAN: Add support for few more statistics to DroneCAN RC by olliw42 · Pull Request #28390 · ArduPilot/ardupilot · GitHub

  • Allow for different types of quality metric for DroneCAN RC in
  • Thomas and Peter to review

UTC2325 https://github.com/ArduPilot/ardupilot/pull/31177

  • Reset height and position from parameters
  • Randy’s working on this now
  • Origin is now set soon after startup
    • Not on arming
    • This now replaces the Sub thing
    • Slightly different
    • Did it on arming like Andy’s did

UTC2331 - common.xml: specify enumeration for MAV_CMD_NAV_VTOL_LAND.param1 by peterbarker · Pull Request #469 · ArduPilot/mavlink · GitHub

  • Add an enum attribute into an xml message
  • Merged!

UTC2332 - https://github.com/ArduPilot/ardupilot/pull/31619

  • Add fence alt_min_tp and alt_max_tp
  • Home and origin being treated as being the same thing, which is wrong
  • Test flown without issue

UTC2344 - Add New Target DAKEFPVH743_SLIM by engineer-dakefpv · Pull Request #31733 · ArduPilot/ardupilot · GitHub

  • New DakeFPVH743-Slim target
  • Merged!
  • Some more discussion over image sizes

UTC2347 - Replace the Attitude Command Model with the improved Real Time S-Curve algorithm by lthall · Pull Request #31849 · ArduPilot/ardupilot · GitHub

  • Replace attitude command model with improved real-time S-Curve
  • Shouldn’t need to init the attitude controller as we’re going into a mode (e.g. poshold to loiter repeatedly and you might start tot drift down-wwind
  • Reviews left

UTC0007 - Random autotest fixes by peterbarker · Pull Request #31899 · ArduPilot/ardupilot · GitHub

  • Random autotest fixes
  • Merged!

UTC0011 - AP_Motors: Add ESC startup delay in ground idle spool state by lthall · Pull Request #31963 · ArduPilot/ardupilot · GitHub

  • Add ESC startup delay in ground idle spool state
  • Some ESCs take a long time to come up from not-spinning to ready-to-take-commands
    • Doing their own spool-up logic
  • This PR adds a time in seconds which can be set so you don’t try to take off while the motors are in this spool-up state
  • Logic really should be in the motors library, which is the subject of a new PR

UTC0021 - AP_Networking: Layer-2 switch architecture for multi-interface networking with switch ports for lwIP, ethernet, serial, and mavlink tunneling by jschall · Pull Request #31966 · ArduPilot/ardupilot · GitHub

  • Layer 2 switch architecture for multi-interface networking in AP_Networking
  • This is changed from what tridge looked
  • Not flown yet
  • Need to know botnet doesn’t break
  • Particularly making sure ppp works
  • An AP_Periph or a vehicle can function as a network switch
  • Can allow for link aggregation
  • 12.5Mbit serial ports
    • Roughly 1MB/second over http
  • No reliance on ppp any more
    • Meaning we can do multicast and broadcast
  • Missing code for ethernet capture
  • LWIP should be able to do some of this stuff?
    • That’s “bridiging”, not really what’s wanted
  • AP_Networking2
    • A lot of scepticism
  • Lots of testing to be done!
  • From the chat:
    • [11:22 AM]Craig.Elder: AP_Networking: Layer-2 switch architecture for multi-interface networking with switch ports for lwIP, ethernet, serial, and mavlink tunneling by jschall · Pull Request #31966 · ArduPilot/ardupilot · GitHub
    • GitHub
    • AP_Networking: Layer-2 switch architecture for multi-interface netw…
    • This change introduces a Layer-2 switch architecture for multi-interface networking in AP_Networking. The switch routes Ethernet frames between pluggable "switch ports" for lwIP (…
    • AP_Networking: Layer-2 switch architecture for multi-interface netw…
    • [11:26 AM]Tim Tuxworth: “should work”???
    • [11:27 AM]rmackay9: Testing with a CubeNodeETH or DroneNet would be best
    • [11:27 AM]rmackay9: Ethernet Adapters and Switches — Copter documentation ← list of supported ethernet switches and adapters
    • [11:28 AM]Tim Tuxworth: Could also be a uart → GPIO on a Raspberry Pi running pppd
    • [11:31 AM]Thomas: i’m still confused why we can’t use the bridge module already in lwIP for this, and the already existing PPP as a protocol
    • [11:31 AM]Thomas: instead of inventing both again from whole cloth
    • [11:32 AM]Thomas: one megabyte per second is still less than 1x12.5Mbps, why can we not fix that up?
    • [11:32 AM]rmackay9: so it does ethernet packet forwarding somewhat like our serial forwarding which is interesting..
    • [11:32 AM]Tim Tuxworth: Why not just use an AP with an ethernet port?
    • [11:34 AM]rmackay9: It would save cost and weight to be able to do ethernet without the switch
    • [11:34 AM]Thomas: at the cost and weight of our code and binary sizse
    • [11:34 AM]Tim Tuxworth: tradeoffs - if you want fast use fast technology.
    • [11:35 AM]rmackay9: The botblox dronenet is quite expensive $120
    • [11:39 AM]Thomas: @Tom Pittenger we can put our heads together later possibly but 802.1D is in the first line of wikipedia about network switches
    • [11:40 AM]Tom Pittenger: I looked into it for LWIP. It didn’t seem like the right fit
    • [11:41 AM]Tom Pittenger: A “network bridge” is definitely what we want, like this
    • Network bridge - Wikipedia
    • Network bridge
    • A network bridge is a computer networking device that creates a single, aggregate network from multiple communication networks or network segments. This function is called network bridging. Bridging is distinct from routing. Routing allows multiple networks to communicate independently and yet remain separate, whereas bridging connects two separ…
    • Network bridge
    • [11:41 AM]Thomas: IEEE 802.1D – Standard which includes bridging, Spanning Tree Protocol and others
    • [11:41 AM]Tom Pittenger: but those require MAC/PHYs. To do withotu that gets more complicated
    • [11:42 AM]Thomas: if you could cast it in terms of existing lwip blocks tridge would surely be much happier too
    • [11:42 AM]rmackay9: I think we just need a dev or two to give the PR a try and we can get an idea of how far we are away from it being merge-able
    • [11:43 AM]rmackay9: focusing on what we already support rather than new features
    • [11:43 AM]rmackay9: I suspect we’re only talking about an hour or so of investigation and back-and-forth with @jschall
    • [11:43 AM]rmackay9: I’m away until next Thu but I’d be happy to give it a go after that
    • [11:43 AM]Tom Pittenger: agreed. It needs to 100% work with existing deployed PPP systems and the params need to be documented and not a “totally new thing”
    • [11:44 AM]Thomas: completely reinventing things is a bad idea!
    • [11:45 AM]rmackay9: I am also a big believer in working with what you’ve already got
    • [11:48 AM]rmackay9: The reason that I’m slightly keen is that removing the requirement for the switch could greatly increase the adoption of ethernet
    • [11:49 AM]rmackay9: ask the average DIY users to pay an extra $120 and they’ll say, “I prefer serial, thanks”
    • [11:49 AM]Peter Hall: I’m not sure it would in most cases.
    • [11:49 AM]Thomas: i think most people do not need a switch
    • [11:49 AM]Peter Hall: I don’t think it would be a good idea to be trying to push camera streams via the flight controller.
    • [11:51 AM]jschall: who says this is doing that?
    • [11:51 AM]Peter Hall: No one, just that is why most have a switch on there vehicle.
    • [11:52 AM]Peter Hall: So this does not remove the need for that switch.
    • [11:52 AM]Peter Hall: You still need a “real” switch.
    • [11:54 AM]Thomas: the list could also help you decide what could be added and where. like there’s all these “cool tricks” but what use cases?
  • Complete rewrites will get push-back!
  • Pep talk from Leoard saying the more criticism you get the better
    • Don’t get despondent
  • Mavlink transport?
  • SITL + TAP
    • Ethernet frames over mavlink!
    • Via your computer just speaking mavlink you get to talk to an ethernet device your system
    • Tunnel message and a type on that
  • “COBS”
    • Arbitrary data going over a serial link
    • Framing for arbitrary protocol
    • Zero byte is always end or start of a frame
    • Byte-stuffs zeroes
    • Framing and then crc is appended
    • Bonding includes sequence number

UTC0110 - Add ability to compute POI as a fallback without Terrain enabled by Hwurzburg · Pull Request #32010 · ArduPilot/ardupilot · GitHub

  • POI without terrain enabled
  • A class variable which can be made local
  • Randy’s happy for this to be merged if Peter reviews and is happy

UTC0115 - Add CI check that added or modified markdown files pass linting check by peterbarker · Pull Request #32014 · ArduPilot/ardupilot · GitHub

  • Add linting of markdown added to tree
  • So new board readmes must be valid
  • Merged!

UTC0118 - AP_HAL_ChibiOS: update AIRBRAINH743 pinout images by andyp1per · Pull Request #32031 · ArduPilot/ardupilot · GitHub

  • Update pinout images
  • MergeOnCIPass

UTC0120 - AP_HAL_ChibiOS: rebuild SequreH743 bootloader for ECC fixes by andyp1per · Pull Request #32032 · ArduPilot/ardupilot · GitHub

  • Testing plan is not complete
  • This is a 4.6 merge
  • What about master?

UTC0124 - Scripting OSD by andyp1per · Pull Request #32045 · ArduPilot/ardupilot · GitHub

  • Scripting OSD
  • Had a good review from Peter

UTC0125 - hwdef: disable GHST telemetry and Video TX for Pixhawk1-1M by andyp1per · Pull Request #32062 · ArduPilot/ardupilot · GitHub

  • Disable GHST and VideoTX for Pixhawk1-1M
  • Many more boards, not jsut Pixhawk1-1M
  • Still available on custom build

UTC0132 - mavlink: move forward to ardupilot/master by peterbarker · Pull Request #32066 · ArduPilot/ardupilot · GitHub

  • mavlink: move forward to ardupilot/master
  • Maybe we can do signature override on PARAM_VALUE_EXT, allowing for the patch in pymavlink which we are reverting
  • Merged!

UTC0136 - hwdef: Update Morakot by Hwurzburg · Pull Request #32067 · ArduPilot/ardupilot · GitHub

  • Update Morakot
  • Comments left
  • Invalid markdown?

UTC0139 - AP_NavEKF3: fix incorrect delta velocity bias covariance inhibit by tpwrules · Pull Request #32069 · ArduPilot/ardupilot · GitHub

  • Fix incorrect delta velocity bias covariance
  • Fixes issue on three replay logs showing a bug
  • Does change replay results generally, ‘though
  • Randy will flight test but then can be merged

UTC0143 - AP_Vehicle: store currently running firmware version into a parameter by peterbarker · Pull Request #32071 · ArduPilot/ardupilot · GitHub

  • Scribble down current version number in a parameter
  • Hidden or internal use?
  • What will this do to people who play with their own version numbers?
    • People don’t just use custom build name
    • They play directly with the version numbers themselves
  • More trouble than this is worth?
  • Use a secret evil version number instead of the version number?
  • Parameter conversion number - any time we remove numbers

12:43 PM]Craig.Elder: AP_Vehicle: store currently running firmware version into a parameter by peterbarker · Pull Request #32071 · ArduPilot/ardupilot · GitHub

GitHub

AP_Vehicle: store currently running firmware version into a paramet…

save the currently running firmware version into a parameter. Into the future we may use this number to prevent a user from attempting to upgrade from a truly ancient version of the firmware to a …

save the currently running firmware version into a parameter. Into the future we may use this number to prevent a user from attempting to upgrade from a truly ancient version of the firmware to a …

[12:45 PM]rmackay9: have we really seen this problem though?

[12:45 PM]Robert L.: I have a basement full of old stuff that would trip over this…

[12:48 PM]Tim Tuxworth: Thomas explained that

[12:49 PM]rmackay9: how about companies that use their own version numbers?

[12:49 PM]Tim Tuxworth: yes you do

[12:51 PM]rmackay9: I think they should be visible if we’re going to pop errors based on the version numbers

[12:51 PM]Robert L.: I just found one heli with V3.3.1 on it. :rofl:

[12:58 PM]rmackay9: let’s rename the project!

[12:59 PM]rmackay9: shouldn’t we wait until we actully have a need for the params?

[12:59 PM]rmackay9: I thought we had a golden rule that we don’t add stuff that isn’t used

[1:00 PM]Thomas: the problem is we would need it three years ago

[1:01 PM]Tim Tuxworth: Oh that’s too much rigmarole

[1:02 PM]rmackay9: I need to go I’m afraid..

[1:02 PM]Tim Tuxworth: New incremental number starting at 1 with 4.7.0

[1:03 PM]Tim Tuxworth: it’s a “build number”

[1:03 PM]Craig.Elder: Copter: skip esc calibration startup check for DSHOT ESCs by rishabsingh3003 · Pull Request #32072 · ArduPilot/ardupilot · GitHub

GitHub

Copter: skip esc calibration startup check for DSHOT ESCs by rishab…

This could potentially waste 2 seconds of startup time

Copter: skip esc calibration startup check for DSHOT ESCs by rishab…

[1:03 PM]Thomas: we are on git we are past that :sob:

[1:04 PM]Tim Tuxworth: it’s a tag

[1:06 PM]Peter Barker: @Robert L. you will already hit problems trying to move your really, really old vehicles to modern ArduPilot as we currently simply delete old parameter conversion code without telling the user that we’ve dropped their parameter values entirely. The code I’m putting forward would tell you that youi’re trying to do something which will not preserve your current behaviour.

UTC0203 - Copter: skip esc calibration startup check for DSHOT ESCs by rishabsingh3003 · Pull Request #32072 · ArduPilot/ardupilot · GitHub

  • Save 2 seconds on boot when using dshot output
  • MergeOnCIPass

UTC0207 - MAVProxy on Windows

  • Sometimes really, really slow
    • Changes between laptops
  • MO does signing of drivers and it is hard to install unsigned drivers
  • If we can fix this it would be a huge improvement for a large number of users

UTC0217 - close