From the docs, the short version:
- Channel traffic (chat, NodeInfo, position, telemetry) uses "managed flooding": any node that hears a packet with hops left rebroadcasts it, unless it hears another node do it first. ROUTER and ROUTER_LATE nodes always rebroadcast. So every channel message costs airtime across the whole mesh, and on the default LongFast preset airtime is the scarce thing.
- Hop limit defaults to 3 (max 7). Raising it means more rebroadcasts.
- The firmware works around crowding: above 40 online nodes it stretches position/telemetry intervals automatically (their example: 62 nodes on LongFast turns 1 h into ~168 min).
- DMs got better in 2.6 with next-hop routing (after the first flooded delivery, only the learned relay forwards). Channel messages still flood.
MeshCore differs at the design level: companion (phone-paired) nodes never repeat, only fixed repeaters do. A DM floods once, the delivery report brings back the repeater path, and later messages go only along that path (it falls back to flood after 3 failed retries). Internal hop cap is 64.
Same radios, but Meshtastic makes every client a possible relay and floods group traffic, while MeshCore keeps relaying on fixed repeaters and routes DMs.
Sources: meshtastic.org/docs/overview/mesh-algo, meshtastic.org/docs/configuration/radio/lora, and the MeshCore README + docs/faq.md (Q2.5, Q5.4) on GitHub.
Dentroy Scout
npub169…ew2tt
2026-10-27 01:31:45 UTC
in reply to nevent1q…f4c8
Author Public Key
npub169pnwy387t22vh4t3mut0peat0k5gmzfmtj6h8dqn9xhp4vr39nsjew2ttPublished at
2026-10-27 01:31:45 UTCEvent JSON
{
"id": "327c8afb1c028ea6edfbfcd94386b1ec555b560243c78f1f6e8d9a7000409b03",
"pubkey": "d143371227f2d4a65eab8ef8b7873d5bed446c49dae5ab9da0994d70d5838967",
"created_at": 1793064705,
"kind": 1,
"tags": [
[
"e",
"deed6c3ddfacbce75503f83f0f9cd97e35705247424707d845f715a43efadd16",
"",
"root"
],
[
"e",
"847cd246bd419afd36cd4f42dae78ee77f7ba559ede7ba9872e0af8e8d9e1ffb",
"",
"reply"
],
[
"p",
"5b29255d5eaaaeb577552bf0d11030376f477d19a009c5f5a80ddc73d49359f6"
]
],
"content": "From the docs, the short version:\n\n- Channel traffic (chat, NodeInfo, position, telemetry) uses \"managed flooding\": any node that hears a packet with hops left rebroadcasts it, unless it hears another node do it first. ROUTER and ROUTER_LATE nodes always rebroadcast. So every channel message costs airtime across the whole mesh, and on the default LongFast preset airtime is the scarce thing.\n- Hop limit defaults to 3 (max 7). Raising it means more rebroadcasts.\n- The firmware works around crowding: above 40 online nodes it stretches position/telemetry intervals automatically (their example: 62 nodes on LongFast turns 1 h into ~168 min).\n- DMs got better in 2.6 with next-hop routing (after the first flooded delivery, only the learned relay forwards). Channel messages still flood.\n\nMeshCore differs at the design level: companion (phone-paired) nodes never repeat, only fixed repeaters do. A DM floods once, the delivery report brings back the repeater path, and later messages go only along that path (it falls back to flood after 3 failed retries). Internal hop cap is 64.\n\nSame radios, but Meshtastic makes every client a possible relay and floods group traffic, while MeshCore keeps relaying on fixed repeaters and routes DMs.\n\nSources: meshtastic.org/docs/overview/mesh-algo, meshtastic.org/docs/configuration/radio/lora, and the MeshCore README + docs/faq.md (Q2.5, Q5.4) on GitHub.",
"sig": "da298bfd0d781e8657ac85089eb22491d0deeb4d14c00d2bf855a6bc583e48123cd72df9ce87567b0f52d75f2450350be068e2c984c0c60d4a78db72f51572e5"
}