[Feature request] Persistent outbox in the companion: retry direct messages until ACKed
Maintainers usually reply within 1 day
Nobody has claimed this yet.
Assessment
- Difficulty
- 5/5
- Estimated time
- Over a week
- Newbie friendliness
- 25/100
- Issue type
- Feature
- Clarity
- Mostly clear
- Activity status
- Active
- Tech stack
- c
- Domain
- networking
Research direction
Start by reading the related issues #3518, #1489, #1342, and #2974, plus the decision in #613, to understand the existing ACK, retransmission, and TX lifecycle behavior. Then locate the companion firmware’s message handling and companion protocol push code. Done means persistent retry and expiry behavior, deduplication, and outbox state updates work across phone disconnections and intermittent multi-hop reachability.
Written by the indexing model from the issue text.
Description
Goal: a DM is either delivered and ACKed, or reported as expired after a configurable retention time, also when the phone is disconnected and when the recipient is only intermittently reachable over many hops (we measured 4-hop round trips at ~32 km working intermittently, and paths of 8 to 12 hops at ~100 km).
Proposal (endpoint-only, no storage on intermediate nodes, in line with the decision in #613):
- The companion keeps unACKed DMs in a persistent outbox in flash, instead of relying on the app to retry while connected.
- Retry policy in the firmware: direct path first, flood with the configured scope after N failures, an immediate retry when an advert or path update from the recipient is received, and backoff otherwise (e.g. 30 s, 2 min, 10 min, 30 min, then hourly) until the retention time expires.
- The recipient dedups on the existing message identity (timestamp + sender), so a lost ACK and a retransmission do not produce a duplicate in the app.
- New companion-protocol push codes for outbox state changes (queued, sent, ACKed, expired), so apps can show delivery state even for messages that were ACKed while the phone was disconnected.
Why in the firmware and not the app: the radio is the always-on device and hears the recipient's adverts. App-side retries stop when the phone sleeps (notably iOS background limits) or disconnects.
Airtime: retries are mostly presence-triggered and backed off, so the flood load should stay below what users generate today by resending manually.
Depends on / related to: #3518 (ACK sent for a DM that is then dropped), #1489 (ACK reliability), #1342 (retransmission of direct messages), #2974 (TX lifecycle signalling).
Happy to discuss the design and help with smaller code changes, but I can't commit to building the full feature.
- Dominant language
- C
- Stars
- 3.9k
- Forks
- 1.5k
- Avg merge
- 1d 1h
- Merged PRs (30d)
- 26
Getting set up
This project ships no dev container, Dockerfile or contributing guide, so setting up is up to you: start from its README, and see our first-contribution guide for the general steps.
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from meshcore-dev/MeshCore
-
Difficulty 2/5 1-3 hours Newbie friendliness 63/100
meshcore-dev/MeshCore#3622 ·
Maintainers usually reply within 1 day
-
[Feature Request] Set Australia (Mid) Preset to automatically enable 2 byte default path hash sizeOpen
Difficulty 2/5 1-3 hours Newbie friendliness 62/100
meshcore-dev/MeshCore#3571 · 5 reactions ·
Maintainers usually reply within 1 day
-
Difficulty 2/5 1-3 hours Newbie friendliness 74/100
meshcore-dev/MeshCore#3567 · 1 comment ·
Maintainers usually reply within 1 day
-
RTC_RX8130CE writes and reads the one-hot WEEK register as a numberPossibly taken A pull request linked to this issue is open or already merged. Open
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
meshcore-dev/MeshCore#3549 ·
Maintainers usually reply within 1 day
-
Repeater: CLI reply to 'setperm <key> 0' is lost when the removed ACL entry sits before the requester (stale ClientInfo* after array shift)Possibly taken A pull request linked to this issue is open or already merged. Open
Difficulty 2/5 1-3 hours Newbie friendliness 77/100
meshcore-dev/MeshCore#3513 ·
Maintainers usually reply within 1 day
All issues in meshcore-dev/MeshCore
Similar issues
-
Difficulty 2/5 1-3 hours Newbie friendliness 66/100
Maintainers usually reply within 1 day
-
Bug Status: Needs Triage
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
Maintainers usually reply within 1 day
-
interpreter-core type-feature
Difficulty 1/5 Under an hour Newbie friendliness 78/100
Maintainers usually reply within 1 day
-
category:port-update
Difficulty 2/5 1-3 hours Newbie friendliness 68/100
Maintainers usually reply within 2 days
-
Update OpenBSD CI to OpenBSD 8.0 and vmactions/openbsd-vm v1.4.9Possibly taken @hyoklee claimed this today. Open
Difficulty 1/5 1-3 hours Newbie friendliness 85/100
Maintainers usually reply within 2 days