v0.3.2

Date: 2026-10-31
Previous release: v0.3.1

Changed

• Policy 1.2: build output in bin/ and release/, a per-package coverage floor, and more tests and CI. No change in behaviour.
• Built on go-canbus v0.7.2, go-canopen v0.3.3, go-serial v0.2.1.

v0.3.1

Date: 2026-10-31
Previous release: v0.3.0

Dependencies

• go-canopen v0.3.2: program download, scan and LSS node-ID assignment fixes (see its release notes), and go-canbus v0.7.1. No command changes.

v0.3.0

Date: 2026-10-30
Previous release: v0.2.1

Exit codes converge on the organisation table (ADR-0006). Scripts that test for these codes must change.

Case	Old	New
Missing privilege (permission)	7	9
Interrupted before a result (code canceled)	10 (internal)	130 (interrupted)

All other codes are unchanged, 11 (abort) and 12 (device) included. error-v1.schema.json adds the kind interrupted. Exit 7 is now reserved for a partial result, which canopenctl does not produce today.

CANOPENCTL_OUTPUT_FORMAT is now accepted besides CANOPENCTL_FORMAT, which is the canonical name and wins when both are set.

v0.2.1

Date: 2026-10-30
Previous release: v0.2.0

Built on go-canopen v0.3.1 (was v0.3.0).

• boot: go-canopen v0.3.1 fixes a race of the network boot-up. A
  slave whose boot-up arrived before the manager's NMT reset communication
  was already being configured when the reset undid its configuration, and
  boot reported it failed (status K); the boot-up now cancels the running
  boot slave processes before the reset.
• Interop scenarios for the v0.2.0 commands against canopen-interop
  v0.2.0 (contract 1.1, images pinned by digest):
  • lss against CANopenNode's and Lely's LSS slaves: identify, select,
    inquire, bit timing, configure node-ID, store, Fastscan of unconfigured
    and configured devices, assign-all; and the LSS masters of all three
    stacks against simulate --lss (inquire, configure node-ID, store kept
    across a restart with --store, Fastscan, assign-all).
  • mpdo configure, dispatcher, scanner on Lely devices, checked
    with Lely's DAM- and SAM-MPDOs and by SDO; mpdo watch of Lely's MPDOs.
  • boot of every stack's device (concise DCF, identity, start), with the
    other product's identity (error status M, exit 1), with --lss for the
    stacks' LSS slaves, and of a network with a device of every stack;
    Lely's boot master booting simulate.
  • cdcf convert byte for byte against Lely's dcfgen and dictionary dump;
    cdcf apply to every stack's device, read back.
  • store/restore, program and --fd: recorded skips — no device of
    the contract has 0x1010/0x1011 or 0x1F50/0x1F51 (probed: abort
    0x06020000), no stack implements CANopen FD.
• Four new known differences, observed on the bus: CANopenNode v4.1's LSS
  slave answers neither identify service; Lely's LSS slave applies a
  configured node-ID only at NMT reset communication and answers Fastscan
  also when configured. LIMITATIONS.md says how to work with them.

Verification:

• Interop on Linux 7.0 arm64 (vcan0), full suite twice with identical
  results, the new scenarios three more times: 152 pass, 5 known
  differences, 63 skips, 0 failures (before: 93 pass, 1 known, 17 skips).
  Per adapter: canopennode 48/2/23, lely 62/3/8, pythoncanopen 41/0/32,
  mixed network 1 pass.
• make check (fmt, staticcheck, golangci-lint, vet, cross, govulncheck,
  tests with the race detector, coverage 93.0 %); golangci-lint and go vet
  with the interop build tag.

v0.2.0

Date: 2026-10-29
Previous release: v0.1.1

Built on go-canopen v0.3.0 (was v0.2.0): LSS, MPDO, CiA 302 network
management and CANopen FD.

• lss: an LSS master (CiA 305). switch (global), select
  (selective, by LSS address), inquire (address and node-ID),
  configure (node-ID, bit rate from the CiA 301 table, store), activate
  (bit timing switch), identify (non-configured devices, or an address
  range), fastscan (the addresses of every unconfigured device) and
  assign-all (a node-ID for each, with --avoid and --store). A
  refusal is exit 12 (lss_refused), no answer exit 5 (no_response).
• mpdo: configure makes a TPDO or RPDO an MPDO in source or
  destination address mode, scanner and dispatcher write the object
  scanner and dispatcher lists, watch decodes MPDOs (passive).
• boot: the CiA 302-2 network boot-up as a manager — slaves with their
  DCF or concise DCF, expected identities, mandatory and optional slaves,
  --lss for the LSS step that gives unconfigured devices the node-ID of
  the slave whose identity they have, --store, --no-start. The result
  lists each slave's CiA 302 boot status; a failed slave is exit 1.
• cdcf convert, show, apply: concise DCFs (0x1F22) from DCFs
  (PDOs in the CiA 301 sequence), listed, and written entry by entry.
• store and restore (0x1010/0x1011, --group, restore --reset); program download (CiA 302-3: stop, clear, image, flash
  status, identification, start), program control and program status.
• CANopen FD: --fd (and --brs) on sdo read/write/dump (USDO
  client, --block for bulk downloads, --client-id), scan, nmt,
  heartbeat watch, emcy watch (20-byte FD emergencies), sync, pdo show/map, dcf apply and the new commands. monitor --fd and decode
  now decode CAN FD frames: USDO requests and responses (--service usdo)
  and FD emergencies; before, they skipped CAN FD frames. A USDO abort is
  exit 11 with its SDO equivalent and usdoCode.
• simulate: --lss (an LSS slave; without --node-id it starts
  unconfigured), --serial, --fd, --store FILE (store and restore,
  LSS store) and --program (program download objects with an in-memory
  program).
• JSON: new result and record shapes in the schemas (additions only,
  schemaVersion 1.0); new error codes lss_refused, lss_conflict,
  no_node_id, program_failed, boot_failed, mpdo_inactive.

Verification:

• In-process tests of every new command against go-canopen nodes on the
  virtual bus with the race detector: LSS slaves (unconfigured and
  configured, with and without a store), MPDO producer and consumer,
  store/restore and program download devices, the CiA 302 boot-up with
  and without the LSS step, CANopen FD nodes over USDO; JSON output
  validated against the schemas; whole-module coverage ≥ 90 %.
• End to end on vcan0 (Linux 7.0, CAN FD MTU): two unconfigured
  simulate --lss processes found by lss fastscan and booted by boot --lss; a simulate --fd process operated over USDO (sdo read, nmt --wait, program download, scan) while monitor --fd decoded the
  USDO traffic.
• Interop: the existing suite against CANopenNode, Lely CANopen and
  python canopen passes on the canopen-interop v0.2.0 images (pinned by
  digest; contract 1.1 only adds things) in CI on amd64 and arm64. The
  new commands have no interop scenarios yet.

v0.1.1

Date: 2026-10-29
Previous release: v0.1.0

Built on go-canopen v0.2.0 (was v0.1.0).

• sdo read and sdo write with --block now handle empty values. A block
  transfer of 0 bytes may end without a data segment: Lely's SDO server sends
  such an upload, python canopen's client such a download. go-canopen v0.1.0
  read that end message as an out-of-order segment and timed out; v0.2.0
  accepts it.
• The interop suite pins the canopen-interop v0.1.0 images by digest.

v0.1.0

Date: 2026-10-29
Previous release: none

Summary

First release of canopenctl, a command-line toolkit for CANopen networks
built on go-canopen and go-canbus v0.7.0: monitor and decode traffic, SDO,
NMT, heartbeat, EMCY, SYNC, TIME, PDOs, EDS/DCF files, CiA 402 drives and
simulated nodes, on SocketCAN, serial-line CAN adapters and an in-process
virtual bus.

Additions

• monitor decodes live traffic as CANopen messages (NMT, SYNC, TIME,
  EMCY with the meaning of its code, PDOs, SDO with segmented and block
  transfers attributed to their object, heartbeat, boot-up, node guarding,
  LSS), filtered by --node and --service; decode does the same for
  trace files (candump, JSON Lines, pcap, pcapng, ASC, BLF, TRC).
• sdo read, sdo write, sdo dump: expedited, segmented and block
  transfer, data types from --type or --eds, files for DOMAIN objects,
  --verify; an abort code is reported with its meaning and as exit code
  11 (JSON: the error's abort object).
• nmt to one node or all, with --wait for the heartbeat or boot-up
  that confirms it; heartbeat watch (boot-ups, state changes,
  timeouts); scan (device type, profile and identity of nodes 1..127).
• emcy watch, sync (period, count, counter) and time.
• pdo show, pdo map (CiA 301 configuration sequence, checked
  against --eds) and pdo watch (decoded values; passive with
  --cob-id/--map).
• eds validate, eds show, eds convert (EDS ↔ DCF with the
  node-ID), eds generate (typed Go code), dcf apply (with
  --dry-run) and dcf export (a node's values into a DCF).
• cia402 state, enable, disable, fault-reset, quick-stop,
  mode, move, velocity: a drive operated through its statusword,
  each step verified; a drive fault is exit 12 with the drive's state.
• simulate: a complete go-canopen node from an EDS or DCF (or a
  built-in device), optionally a CiA 402 drive, with ticking objects and an
  emergency on request, for demos and tests.
• Safety: every state-changing command needs --yes on an interface
  that is not virtual; passive commands open slcan adapters listen-only;
  productions are bounded.
• --format text|table|json with JSON Schemas (schemas/), JSON Lines
  streams, stable exit codes (docs/exitcodes.md).

Verification

• In-process tests of every command against go-canopen nodes (CiA 401 I/O
  module, built-in device, CiA 402 drive) on the virtual bus, with the race
  detector; golden JSON validated against the schemas; ≥ 90 % coverage.
• End to end on vcan0 (Linux 7.0): simulated node and drive processes
  operated by canopenctl processes.
• Interop on vcan0 against CANopenNode v4.1, Lely CANopen 2.3.5 and python
  canopen 2.4.1 (canopen-interop): canopenctl against their devices, their
  clients against canopenctl simulate, their EDS checkers against
  eds validate.