Write down one interface standard for moving amateur station information between APRS and MeshCore, so that:
In scope (candidates, to be confirmed: OPEN-QUESTIONS Q2):
Out of scope (unless decided otherwise):
| Phase | Output | Done when |
|---|---|---|
| 0. Groundwork | research/ notes: prior art, the MeshCore packet facts, the APRS facts, regulatory notes |
Every spec section has the facts it needs, with sources |
| 1. Architecture | Spec §3-4: roles (gateway, iGate, digi-style bridge), directions, trust model; decision records for each | Q2-Q8 answered |
| 2. Mappings | Spec §5-9: position, identity, messaging, telemetry, objects; field-by-field tables | Each mapping has a table plus test vectors |
| 3. Gateway rules | Spec §10-11: dedup, loops, rate limits, failure behaviour, logging | Rules written as MUST/SHOULD, with vectors for every drop reason |
| 4. Reference implementation | Kelly Page iGate brought to the spec (js/igate.js, bridge/kelly_igate.py) |
Kelly passes every vector |
| 5. Field trial | On-air trial on the SEQ mesh plus APRS-IS; a report | Agreed metrics met (Q12) |
| 6. Publish | Spec 1.0, offered to MBUG and others | Alec signs off |
Phase 4 can start for the parts already settled. The current Kelly iGate is effectively a “version 0” of the position profile.
Alec decides; nothing in the spec is settled until it has a decision record. Outside review is held on the published 1.0 Candidate (decision 0020): MBUG (channel and SEQ settings), the APRS-IS / aprs2.net operators (tocall and q-construct) and the MeshCore developers (data type).
inject.json, text.json and stateful vectors (Phase 4)