Skip to content

Packet: BBS & NET/ROM node

This page turns the ingest box into a connected-mode packet station: a NET/ROM node, a BBS, and FBB mail forwarding with the wider packet network. It is for the sysop; at the end the node answers connects and trades routes with its neighbours.

These services run on the operator's own ingest box, over its radio (the KISS TNC, else the first soundcard port), or over an AXUDP link to other nodes when the box has no radio.

These services transmit automatically

Over a radio, the NET/ROM node, the BBS and FBB forwarding answer and send on the air without an operator at the key. Read Automatic stations on the air before you turn them on with a radio.

Before you start

  • An ingest box with a working radio: a KISS TNC with PTT or a soundcard port with its PTT; or a two-way AXUDP link (AXUDP_PEERS).
  • A callsign-SSID for each service, for example OE8APR-5 for the node and OE8APR-8 for the BBS (Callsigns and SSIDs).
  • Settings go in the box's settings file (deploy/.env in Docker); restart the ingest after each change.

Connected-mode AX.25

The node and the BBS run on the ingest box's own AX.25 stack, over its radio (the KISS TNC, else the first soundcard port) or an AXUDP port. Defaults: t1 3 s, t3 30 s, n2 10, window 4, modulo 8. The modulus is chosen for an outgoing connect and adopted from the peer on an incoming one, and SREJ is opt-in per link. How the stack works is in The AX.25 stack.

Players' tools run in their browsers, never on the ingest box, so they take no part in the node's or the BBS's sessions: those answer with their own commands only. Tools answer on a player's own packet terminal (Tools on connected sessions).

NET/ROM node

Set NETROM_CALL and NETROM_ALIAS (both required) to run a node over the box's radio or, without one, over a two-way AXUDP port. The node:

  • broadcasts its NODES table every NETROM_BROADCAST_MS (default 3600000, one hour; at least five minutes) and learns routes from the NODES broadcasts it hears, letting stale routes age out; NETROM_PATH_QUALITY (0–255) is the quality it assumes for a neighbour heard directly;
  • switches directed NET/ROM frames: delivers them locally, forwards them on, or drops them;
  • accepts L4 circuits that end at the node (bound to the node command line), and connect-through (C <dest>), which routes a caller onward and bridges the two;
  • answers with the command surface set by NODE_PERSONALITY: netrom, flexnet, tnn or baycom;
  • mirrors its learned NODES table and its MHeard list to the gateway (/api/node/nodes, /api/node/mheard).

Check that it worked. The box logs [netrom] node <ALIAS>:<CALL> active on <port> and, once it answers connects, [netrom] node CLI answering inbound connects on <CALL>. Learned routes log as [netrom] learned … route(s) and appear in the NET/ROM node app in the Shack and under Instance admin.

INP3 routing

Set NETROM_INP3=1 to also speak INP3 (Improved NET/ROM) beside the NODES broadcasts. INP3 replaces the 0–255 quality and the fixed five-minute flood with triggered, point-to-point Routing Information Frames (RIFs), ranked by the measured round-trip transport time (tt). The network converges faster and prefers paths with low latency. The node:

  • learns routes from a neighbour's directed RIF, adds the link's own tt to each route, and advertises again only what changed, within a 30-hop horizon;
  • measures each neighbour's latency with L3RTT probes and takes a route's tt as half the smoothed round trip;
  • withdraws a route (tt = 60000) when its neighbour does, or when it ages out;
  • shows INP3 routes in the same node table, with their tt mapped to a NET/ROM-style quality for display only; routing stays on tt.

The log shows [inp3] enabled — RIF learning, L3RTT probing, triggered updates on <port>. Classic NODES broadcasting keeps running, so an INP3 node still works with plain NET/ROM neighbours.

BBS

Set BBS_NODE_CALL, and the box answers inbound AX.25 connects to that call with the F6FBB command set (list, read, send, kill, help, …), backed by a per-caller snapshot of the gateway's mail store. The log shows [bbs] BBS answering inbound connects on <CALL> (FBB forwarding gate armed).

The gateway holds the BBS's message base: personal mail, bulletins and NTS traffic. It moves them the F6FBB way only: stations read and write over a connect to the packet BBS, and FBB forwarding exchanges them with partner BBSes. The BBS never sends over APRS or MeshCom. The message format (P/B/T type and BID) is MBL/FBB-compatible. A message posted here gets the BID <number>_<call>, the number in base 36 and the call the sysop's base call, within the 12 characters F6FBB accepts. FBB forwarding therefore needs the gateway's ADMIN_CALLSIGNS (or SERVICE_CALL) to name the BBS's call; the box warns at start when they differ. The public BBS endpoints are in the API reference.

FBB forwarding

Set BBS_FORWARD=1 and BBS_FORWARD_CALL to forward mail. It needs the box's radio (the KISS TNC or a soundcard port) or a two-way AXUDP port. The forwarder opens connected-mode AX.25 sessions to partner BBSes and exchanges mail with the ASCII FBB protocol. It uses hierarchical TO@BBS.#REGION.STATE.CC.CONT addressing, longest-prefix routing, and BID/MID de-duplication, and runs multi-hop connect scripts. It checks its queue every BBS_FORWARD_POLL_MS (default 60000) and presents BBS_FORWARD_SID to partners. The log shows [forward] FBB forwarding scheduler active as <CALL>.

The forwarder offers partners the mail of senders whose base call is control-verified on this instance. Mail from an unverified sender stays on this BBS for its addressee. Federation records go only to a partner marked for federation, and only with FED_BBS on: Federation over FBB, experimental and off by default.

Partners and routing rules are set under Instance admin → FBB forwarding (Instance admin at a glance).

Compressed forwarding. Set BBS_FORWARD_COMPRESS=1 to offer FBB binary compressed forwarding (LZHUF, B1). The binary blocks (SOH/STX/EOT, FA proposals, FS !offset resume) travel in the same forwarding session, and FBB MD5 link authentication is supported. Compression starts only when the partner's SID also advertises the B flag. Against an ASCII-only partner the session falls back to plain ASCII, so the option is always safe to turn on.

Federation sync

The box can carry federation records over the same stack. With FED_LINK_SERVE=1 it answers pull requests on FED_LINK_CALL, and the node answers the FED command; the pages come from your own gateway. With FED_LINK_PULL=1 it dials the peers that publish an ax25 or netrom endpoint, one session per FED_LINK_PULL_MS (default one hour), and hands every page to your gateway, which checks each record's signature. The log shows [fedlink] federation sync answering inbound connects on <CALL> and [fedlink] packet pull active as <CALL>. Both sides, the timings and the limits are in Packet circuit.

A NET/ROM node and FBB forwarding can run over the internet instead of RF, or beside it, on the two-way AXUDP or AXIP ports: set AXUDP_PEERS or AXIP_PEERS (Transports). Your node then joins the wider BPQ-style packet mesh without a radio path to every neighbour. Between two 44Net addresses, see AXUDP and AXIP peering over 44Net. Frames that arrive this way stay Tier C.

Next