ArduPilot EU Dev Call 2026-02-25

Georacer · February 25, 2026, 7:52am

Attendees (unique): 5

UTC0700

github.com/ArduPilot/ardupilot

Add a separate Configure class for Ublox CFG_VALGET/VALSET/VALDEL system

master ← bugobliterator:pr-ubx-cfgv2

opened 08:10AM - 01 Oct 25 UTC

bugobliterator

+2798 -36

Tested units: - [x] M9 - [x] M10 - [x] NEO-F9 - [x] ZED-F9 - [x] F10 - [x]… X20/F20, ublox mentioned they are the same hardware just All bands vs Dual bands support. Features: - [x] Configure GNSS constelations - [x] Configure Signals - [x] Show available GNSS constelation and Signals (GPS1_GNSS_AVAIL and GPS1_SIG_AVAIL) - [x] Common configurations such as common message rate, navfilter, min elev, measurement rate - [x] RF data logging - [x] Raw message logging - [x] Add option to clear all configurations that are not explicitly set through ardupilot, ensures there are no inconsistent configurations - [x] Add method to override configuration through other sources (will be implemented in future PRs, user will be able to set configurations through SDCard, MAVFTP and maybe even through dronecan File process) - [x] Moving Baseline support - [ ] Update lag setting To do complete Reset GPS config: `GPS_AUTO_CONFIG 3` To force new config mechanism for all GNSS units: `GPS_DRV_OPTIONS (bit 8)` _Note that forcing new config mechanism will be deprecated in next major release as this code spends enough time in the wild. Eventually we will force all supported units to use the new config mechanism only_. X20P log: [2026-01-20 15-29-30.bin.zip](https://github.com/user-attachments/files/24755104/2026-01-20.15-29-30.bin.zip) <img width="1004" height="559" alt="image" src="https://github.com/user-attachments/assets/60f4b80f-197c-4cd5-9285-c2a3f2bdf637" />

github.com/ArduPilot/ardupilot

Cleanup PR for #31217 Add a separate Configure class for Ublox CFG_VALGET/VALSET/VALDEL system

master ← bugobliterator:pr-ubx-cfgv2-clean

opened 02:49AM - 25 Feb 26 UTC

bugobliterator

+2775 -271

This PR aims to remove the duplicate code in here #31217 , and I believe is manu…ally reviewable. Confirming that the CONFIG KEYS match: * To check the key values actually match the old I asked claude to write me a small script https://gist.github.com/bugobliterator/a1687bec71adbe4fca883c691e43bb25 * Size check with elf diff shows that the config lists match in size as well. Results: ``` OK CFG_TMODE_MODE = 0x20030001 OK CFG_RATE_MEAS = 0x30210001 OK CFG_UART1_BAUDRATE = 0x40520001 OK CFG_UART1_ENABLED = 0x10520005 OK CFG_UART1INPROT_UBX = 0x10730001 OK CFG_UART1INPROT_NMEA = 0x10730002 OK CFG_UART1INPROT_RTCM3X = 0x10730004 OK CFG_UART1OUTPROT_UBX = 0x10740001 OK CFG_UART1OUTPROT_NMEA = 0x10740002 OK CFG_UART1OUTPROT_RTCM3X = 0x10740004 OK CFG_UART2_BAUDRATE = 0x40530001 OK CFG_UART2_ENABLED = 0x10530005 OK CFG_UART2INPROT_UBX = 0x10750001 OK CFG_UART2INPROT_NMEA = 0x10750002 OK CFG_UART2INPROT_RTCM3X = 0x10750004 OK CFG_UART2OUTPROT_UBX = 0x10760001 OK CFG_UART2OUTPROT_NMEA = 0x10760002 OK CFG_UART2OUTPROT_RTCM3X = 0x10760004 OK~ MSGOUT_RTCM_3X_TYPE4072_0_UART1 = 0x209102FF (partial match: CFG_MSGOUT_RTCM_3X_TYPE4072_0_UART1) OK~ MSGOUT_RTCM_3X_TYPE4072_1_UART1 = 0x20910382 (partial match: CFG_MSGOUT_RTCM_3X_TYPE4072_1_UART1) OK~ MSGOUT_RTCM_3X_TYPE1077_UART1 = 0x209102CD (partial match: CFG_MSGOUT_RTCM_3X_TYPE1077_UART1) OK~ MSGOUT_RTCM_3X_TYPE1087_UART1 = 0x209102D2 (partial match: CFG_MSGOUT_RTCM_3X_TYPE1087_UART1) OK~ MSGOUT_RTCM_3X_TYPE1097_UART1 = 0x20910319 (partial match: CFG_MSGOUT_RTCM_3X_TYPE1097_UART1) OK~ MSGOUT_RTCM_3X_TYPE1127_UART1 = 0x209102D7 (partial match: CFG_MSGOUT_RTCM_3X_TYPE1127_UART1) OK~ MSGOUT_RTCM_3X_TYPE1230_UART1 = 0x20910304 (partial match: CFG_MSGOUT_RTCM_3X_TYPE1230_UART1) OK~ MSGOUT_UBX_NAV_RELPOSNED_UART1 = 0x2091008E (partial match: CFG_MSGOUT_UBX_NAV_RELPOSNED_UART1) OK~ MSGOUT_RTCM_3X_TYPE4072_0_UART2 = 0x20910300 (partial match: CFG_MSGOUT_RTCM_3X_TYPE4072_0_UART2) OK~ MSGOUT_RTCM_3X_TYPE4072_1_UART2 = 0x20910383 (partial match: CFG_MSGOUT_RTCM_3X_TYPE4072_1_UART2) OK~ MSGOUT_RTCM_3X_TYPE1077_UART2 = 0x209102CE (partial match: CFG_MSGOUT_RTCM_3X_TYPE1077_UART2) OK~ MSGOUT_RTCM_3X_TYPE1087_UART2 = 0x209102D3 (partial match: CFG_MSGOUT_RTCM_3X_TYPE1087_UART2) OK~ MSGOUT_RTCM_3X_TYPE1097_UART2 = 0x2091031A (partial match: CFG_MSGOUT_RTCM_3X_TYPE1097_UART2) OK~ MSGOUT_RTCM_3X_TYPE1127_UART2 = 0x209102D8 (partial match: CFG_MSGOUT_RTCM_3X_TYPE1127_UART2) OK~ MSGOUT_RTCM_3X_TYPE1230_UART2 = 0x20910305 (partial match: CFG_MSGOUT_RTCM_3X_TYPE1230_UART2) OK~ MSGOUT_UBX_NAV_RELPOSNED_UART2 = 0x2091008F (partial match: CFG_MSGOUT_UBX_NAV_RELPOSNED_UART2) OK CFG_SIGNAL_GPS_ENA = 0x1031001F OK CFG_SIGNAL_GPS_L1CA_ENA = 0x10310001 OK CFG_SIGNAL_GPS_L2C_ENA = 0x10310003 OK CFG_SIGNAL_GPS_L5_ENA = 0x10310004 OK CFG_SIGNAL_SBAS_ENA = 0x10310020 OK CFG_SIGNAL_SBAS_L1CA_ENA = 0x10310005 OK CFG_SIGNAL_GAL_ENA = 0x10310021 OK CFG_SIGNAL_GAL_E1_ENA = 0x10310007 OK CFG_SIGNAL_GAL_E5A_ENA = 0x10310009 OK CFG_SIGNAL_GAL_E5B_ENA = 0x1031000A OK CFG_SIGNAL_BDS_ENA = 0x10310022 OK CFG_SIGNAL_BDS_B1_ENA = 0x1031000D OK CFG_SIGNAL_BDS_B1C_ENA = 0x1031000F OK CFG_SIGNAL_BDS_B2_ENA = 0x1031000E OK CFG_SIGNAL_BDS_B2A_ENA = 0x10310028 OK CFG_SIGNAL_QZSS_ENA = 0x10310024 OK CFG_SIGNAL_QZSS_L1CA_ENA = 0x10310012 OK CFG_SIGNAL_QZSS_L1S_ENA = 0x10310014 OK CFG_SIGNAL_QZSS_L2C_ENA = 0x10310015 OK CFG_SIGNAL_QZSS_L5_ENA = 0x10310017 OK CFG_SIGNAL_GLO_ENA = 0x10310025 OK CFG_SIGNAL_GLO_L1_ENA = 0x10310018 OK CFG_SIGNAL_GLO_L2_ENA = 0x1031001A OK CFG_SIGNAL_NAVIC_ENA = 0x10310026 OK CFG_SIGNAL_NAVIC_L5_ENA = 0x1031001D OK CFG_SIGNAL_L5_HEALTH_OVRD = 0x10320001 OK CFG_NAVSPG_DYNMODEL = 0x20110021 61/61 matched, 0 issues ```

Andrew: The first PR is a no brainer and should get no-compiler-output.

Then the next one would be much cleaner to inspect and merge.
Would be nice to have it for 4.7.

Randy: We branch out the beta on Friday.
A: There isn’t really time to test this PR in real life.
Peter: Will the X20 have GPS for yaw support?
Sid: Not initially.
A: How about single module dual antenna?
S: Also eventually.
George: What’s the selling point for this module?
S: All bands and constellations available. Some issues with GLONASS.

Also corrections from one or two constellations, some freely available.

UTC0719

A: We would need to remove some rangefinders by default, to make space.
R: The JRE is a good candidate for removal.
P: We need to be careful to break as few existing setups as possible.
R: The TFS20 is small and nice, could be nice to include it.
A: You two could go over this list before 4.7.
P: I’d like to freeze the minimized board features eventually.

UTC0726

github.com/ArduPilot/ardupilot

Tools: allow size_compare_branches.py to test boards with modified hwdefs

master ← peterbarker:pr-claude/size-compare-branches-test-modifies-boards

opened 12:36PM - 09 Feb 26 UTC

peterbarker

+99 -0

With just this commit: ``` ValueError: No modified boards found between pr/si…ze-compare-branches-test-modifies-boards and master ``` Modifying `libraries/AP_HAL_ChibiOS/hwdef/CubeOrange/hwdef.inc` in a test commit on top: ``` pbarker@crun:~/rc/ardupilot-claude(pr-claude/size-compare-branches-test-modifies-boards)$ ./Tools/scripts/size_compare_branches.py --modified-boards SCB: Running (git symbolic-ref --short HEAD) in (.) SCB-GIT: pr-claude/size-compare-branches-test-modifies-boards SCB: Running (git merge-base pr-claude/size-compare-branches-test-modifies-boards master) in (.) SCB-GIT: 92d46eca62c3e380af26737f74a6fa914739a5ea SCB: Running (git diff --name-only 92d46eca62c3e380af26737f74a6fa914739a5ea pr-claude/size-compare-branches-test-modifies-boards) in (.) SCB: Modified hwdef: libraries/AP_HAL_ChibiOS/hwdef/CubeOrange/hwdef.inc SCB: Board CubeOrange-periph-heavy uses modified hwdef SCB: Board CubeOrange-bdshot uses modified hwdef SCB: Board CubeOrangePlus-bdshot uses modified hwdef SCB: Board CubeOrange-ODID uses modified hwdef SCB: Board CubeOrange-periph uses modified hwdef SCB: Board CubeOrange-joey uses modified hwdef SCB: Board CubeOrangePlus-ODID uses modified hwdef SCB: Board CubeOrange-SimOnHardWare uses modified hwdef SCB: Board CubeOrangePlus-SimOnHardWare uses modified hwdef SCB: Board CubeOrange uses modified hwdef SCB: Board CubeOrangePlus uses modified hwdef SCB: Running (git merge-base pr-claude/size-compare-branches-test-modifies-boards master) in (.) SCB-GIT: 92d46eca62c3e380af26737f74a6fa914739a5ea SCB: Using merge base (92d46eca62c3e380af26737f74a6fa914739a5ea) . . . SCB: Running (rsync -ap build/ /tmp/tmpe56a2y80/out-branch-CubeOrangePlus-SimOnHardWare) in (.) Board,plane CubeOrange,0 CubeOrange-ODID,* CubeOrange-bdshot,* CubeOrange-joey,* CubeOrangePlus,* CubeOrangePlus-ODID,* CubeOrangePlus-bdshot,* pbarker@crun:~/rc/ardupilot-claude(pr-claude/size-compare-branches-test-modifies-boards)$ ```

Merged!

UTC0726

github.com/ArduPilot/ardupilot

AP_GPS: correct compilation for disabled UBLOX_RXM_RAW_LOGGING

master ← peterbarker:pr/ubx-raw-logging-fixes

opened 08:56AM - 18 Feb 26 UTC

peterbarker

+4 -2

Tested with: ``` ./waf configure --board=CubeOrangePlus --extra-hwdef=/tmp/ext…ra.hwdef ``` ``` pbarker@crun:~/rc/ardupilot(master)$ cat /tmp/extra.hwdef define UBLOX_RXM_RAW_LOGGING 0 pbarker@crun:~/rc/ardupilot(master)$ ``` Master doesn't compile if you modify the source code to turn the option off. This define will need renaming before we expose it further, really. Should be `AP_GPS_UBLOX_RXM_RAW_LOGGING_ENABLED` . So I'm not adding it to `build_options.py`

P: I’ll close this one and defer to Sid’s PR.

UTC0727

A: We now want SITL to match the higher-end boards.

Approved!

UTC0729

P: Our new PR template got ignored the moment we put it up.

Maybe we should point to it in AGENTS.md as well.

A: Approved and Merged!

UTC0733

github.com/ArduPilot/mavlink

common.xml: add MAV_DO_REPOSITION_FLAGS_RELATIVE_YAW flag

master ← peterbarker:pr/reposition-flags-yaw

opened 07:04AM - 24 Feb 26 UTC

peterbarker

+4 -1

These are patches from mavlink/mavlink/master - these patches were merged there …in https://github.com/mavlink/mavlink/pull/2280 ( @IamPete1 ) We don't support this flag. Any GCS currently sending this message in would just have it completely ignored by any and all versions of ArduPilot. Is that what ArduPilot should by default to doing, or should be reject any command that has an option bit set in it that we do not understand? We can't tell ahead of time whether a vehicle which doesn't obey the bit might behave in a manner which the GCS was not expecting....

P: We don’t support this feature. What should we do?
R: Documentation is the answer here.
P: The GCS can’t do much just from documentation.

Should we try our common.xml to match the upstream?

A: In general yes. But I don’t think we can make them 100% identical.

George: Why not NACK such unsupported commands?
P: It is doable, but it will cost some code and flash.
R: Being 100% correct isn’t the best vs doing the best we can.

Merged!

UTC0741

github.com/ArduPilot/mavlink

common.xml: update description of FLIGHT_INFORMATION message

master ← peterbarker:pr/flight-information-description

opened 01:48PM - 18 Feb 26 UTC

peterbarker

+11 -4

from mavlink/mavlink/master This adds another field, so patches will be requi…red to ArduPilot to set the new field. These changes were approved by @nexton-winjeel in https://github.com/mavlink/mavlink/pull/2067 and he did the ArduPilot implementation - so I think we should be good with our implementation with the exception of the new field.

UTC0744

P: I still need to write a test for this.