20190109 Dev Meeting

Agenda

General:

Update from previous calls
- AI @ALL - review RFC in #861 "Local and body frames" for next meeting.
Tagged: issues, PR. A few choice items:
Pymavlink#237: message CRC checksum is wrong when a mavlink_message_t is converted to a sending buffer - quick update.

HamishW

QGC special handling for Param
- Don wants us to confirm that only messages from component MAV_COMP_ID_AUTOPILOT1 should be allowed to have special non-specification handling by a GCS. Messages with any other source component id will be handled according to the appropriate MAVLink spec.
- FWIW, while in theory MAVLink should say "everything must be to spec", in practice we still need to answer the question for "versions". So I think that we'd be saying something along the lines of:
  
  A GCS may provide support for different versions/variants of a protocol. A GCS is only expected to support versions for messages originating from an autopilot (component ID: MAV_COMP_ID_AUTOPILOT1). Messages from other components should be treated according to the most recent supported version of the specification.

CHECK Status

@JamesP and @MichaelS (or Lorenz/Thomas) to engage with Local and body frames and add comments. Discuss next week.
@WickedShell File format discussion - Attempt to re-engage in next two weeks.
@JulianOes to review and respond to message CRC checksum is wrong when a mavlink_message_t is converted to a sending buffer