20190220 Dev Meeting

Agenda

General:

• Update from 20190206 (relevant previous calls)
• Tagged: issues, PR. A few choice items:

HamishW

• All components expected to broadcast HEARTBEAT when not faulted OR are only drones expected to broadcast, and ground stations expected to listen and publish after they receive (to answer this)
• Meet the Rust Guys - Friday Feb 22 (10am -5pm PST) ?
• Support reserved/default params in generated libraries - JamedP
• RC RSSI prior art
• Doc Mission partial upload/download

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.

Notes

Call 1

• Rust Team want to meet to talk about moving generator into MAVLink project. Lorenz to attend. AI on Hamish to connect them and to see if DanA wants to go.
• HEARTBEAT broadcasting - everything should broadcast all the time OR are only Autopilots expected to publish on defined port and GCS listen and publish when have seen autopilot.
  • JulianO Problem: where should GCS publish - it is listening on port on its own computer, but does not have IP/target to publish to.
  • LM:
    • Publishing is mechanism used for discovery, and in system with multiple GCS we will still need to do discovery.
    • Need to publish, but also may need to define publish address/mechanism.
  • But what about GCS not setting unique ids:
    • Discussion that perhaps we need less manual approach to id allocation - centralised ID allocator, or peer allocator.
  • Action on Julian to find out existing best practice in other networks that might help here and report back.
• Support reserved/default params in generated libraries #269
  • Discussed if there are ways this might be enforced at API level (probably not)
  • Michael withdrew objections following understanding that change is docs only.
• 1083: Update RSSI values to 0-254, 255=not used
  • This looks fine. AI on Hamish to merge and add note to different teams to adopt the update.
• STATUSTEXT_LONG - suggest add a sequence and fragment field to allowed us to construct much longer statustexts if needed "before it is too late".
  • This is part way to an event notification. Perhaps best to add field now.
  • Team agreed this is a good idea. AO for Michael DB to create PR.
• RADIO_SETTINGS
  • Most of the fields are details that you wouldn't want except during testing.
  • Probably not enough fields, so eventually will need to extend
  • Suggestion is to use Param protocol for this. We did similarly for ESP8226 configuration and solution pretty scaleable.
  • Otherwise these should not be common, because too specific to this use case.
  • Invite dev call in two weeks
• Doc Mission partial upload/download
  • Lorenz to review
• #926 - Add new PWM output message
  • Lots of discussions around how to mark not used channels, extensiblity, ints vs floats, truncation.
  • AI Lorenz to update to use a bitmask for marking active channels, and to allocate 32.
• #995 - WIP: Added fields for RTL and LAND time estimates
  • Great idea - definitely needed
    • There is a known subset of things we might need, but variation in them. i.e. FW Land can be done in lots of ways, and time depends on what was selected.
    • Would like to define a "standard set"
    • AI Lorenz (Done) to invite author to discuss with us at next dev meeting.

AI:

• JulianO: Report back on possible System ID allocation approaches and GCS discovery mechanisms/publishing heartbeat when not connected.
• DONE HamishW: Connect Rust team (Michal) to Lorenz for meeting.
• DONE (ArduPilot#10584, PX4#11514) HamishW: Merge 1083: Update RSSI values to 0-254, 255=not used and notify JamesP in AP issue.
• MichaelDb: PR to update STATUSTEXT_LONG with fields to allow longer messages.
• DONE HamishW - Respond to RADIO_SETTINGS with invite to dev call and reasoning.

Attendees

HamishW, LorenzM, JamesP, JulianO, Michael DB