20230621 Dev Meeting

General:

• Update from relevant previous calls
• Tagged: issues, PR.

Notes

• Mission_checksum

• A bit of general discussion on alternatives/why this can't work

• AO Hamish to create PR https://github.com/mavlink/mavlink/pull/2012

• Add Payload and Autonomy Microservice Support from Marine Dialect #1998

  • As a dialect it can be added, but will need ranges, less generic name than marine, and clear docs to indicate that it is a dialect
  • Seb didn't turn up to discuss. But we quite like aspects of the proposal.
  • AO Hamish to ping. - Done https://github.com/mavlink/mavlink/pull/1998#issuecomment-1601907570

• RADIO_LINK_STATS message #1920

  • Dronecan co-ordination - https://github.com/dronecan/DSDL/pull/33
  • Basically MAVLink OK as long as they include the information that we want so we can take a subset. It looks like they do, so MAVLink are OK with their messages.
  • Naming conventions are inconsistent between the two - uplink downlink vs RX, TX.

• XSD - can we get these merged?

  • mavschema.xsd - max, min, increment on field #811

  • mavschema.xsd: add superseded tag #809 #809

  • XSD Schema - there can be at most 7 params #814

  • #821 XSD: MAV_CMD missionOnly attribute

  • Add altitudeOnly attribute #824 (part of MAV_CMD_DO_GO_AROUND - isAltitude attribute discussion #1994

  • AO on James again.

• Need a better way to call python generator - from Pymavlink - get answer?

Updates

• Replace validator with xmlschema (XSD1.1 compliant) #816
  • Note, not done.
• Added "shortest way" as an option for yaw direction in MAV_CONDITION_YAW #1962
  • AO James to get Peter to look
• mavlink_msg_xxx_decode by mavgen_c does not handle extensions of message properly #684 - pymavlink bug fix
  • AO JamesP to see if we can get review from PeterB
• message_definitions: add all_with_development.xml #1924
  • JamesP seeking more suggestions.
• 1978 Wireshark mavlink_2_common.lua:9330: bad argument #1 to 'rshift' (number expected, got nil)
  • Some progress FYI - Roku (Dan Rose) fixed that bug but is looking at other problems with output.

Attendees: Hamish, JamesP, JulianO