Skip to content

Rename v2.3 draft to stable filename; settle numbering at 2.3.0; adopt git tags - #75

Merged
M0LTE merged 3 commits into
mainfrom
rename-v23-draft
Jul 16, 2026
Merged

M0LTE merged 3 commits into
mainfrom
rename-v23-draft

Conversation

@M0LTE

@M0LTE M0LTE commented Jul 16, 2026 •

Copy link
Copy Markdown
Collaborator

Summary

  • doc/ax.25.2.3.1_May_26.md → doc/ax.25.2.3-draft.md (git detects the rename at 98% similarity, so git log --follow and blame carry through).
  • Numbering settled at v2.3.0: the draft's title, front-matter revision entries and tag scheme move from Bill's 2.3.1 to 2.3.0 — the working group's first v2.3 release will be v2.3.0. The renumbering is recorded as a new row in the draft's changelog table; the row recording Bill's original re-versioning is kept as written (historical record).
  • Revision tracking via annotated git tags, four-part for draft iterations: ax25-v2.3.0.0, ax25-v2.3.0.1, … with ax25-v2.3.0 reserved for the release itself. Documented in the README and in the draft's changelog-table intro.
  • README link updated — the only reference to the old path in the repo.

Rationale

Date-stamped filenames break inbound links and fragment git history on every revision cut. A stable path plus per-iteration tags gives durable links, clean history, and a checkout-able snapshot of any iteration (git show ax25-v2.3.0.0:doc/ax.25.2.3-draft.md).

Tags are namespaced ax25- because tags are repo-wide and this repo also hosts the KISS, multi-drop KISS, IL2P and FBB documents — leaving room for e.g. il2p-v0.7 later.

Caveats (non-blocking)

  • GitHub does not redirect renamed file paths: the old deep link in groups.io msg 120 will 404 once this merges. Future list posts should link the stable path.
  • Bill's local copies/emails use the old filename and the 2.3.1 numbering — worth a heads-up on the list so the WG is aligned on 2.3.0 before his next cut.

After merge

Push the first tag on the merge commit:

git tag -a ax25-v2.3.0.0 -m "AX.25 v2.3 draft iteration 0: Revision 0 (24 May 2026, renumbered from 2.3.1) plus repo errata to date"
git push origin ax25-v2.3.0.0

(Say the word and I'll do it.)

🤖 Generated with Claude Code

https://claude.ai/code/session_01XodLuT1R8a4HY76RSZrqFi

M0LTE and others added 2 commits July 16, 2026 22:32
doc/ax.25.2.3.1_May_26.md -> doc/ax.25.2.3-draft.md. Date-stamped
filenames break links and fragment history on every revision; the
draft now lives at a stable path, with each revision marked by an
annotated git tag (ax25-v2.3.1, ax25-v2.3.2, ...) matching the
document's own "Version 2.3 Revision N" numbering. Tags are
namespaced ax25- to leave room for tagging the other specs in this
repo. Convention documented in the README and the draft's changelog
intro.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XodLuT1R8a4HY76RSZrqFi
The working group will release v2.3.1 (Bill's chosen number); draft
iterations do not increment that. Tags are ax25-v2.3.1.0, .1, .2, ...
per iteration, with ax25-v2.3.1 reserved for the release itself.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XodLuT1R8a4HY76RSZrqFi
Resolve the numbering question now rather than after tags and list
links accumulate: the working group's first v2.3 release will be
v2.3.0, so the draft title, front-matter revision entries and tag
scheme all move from 2.3.1 to 2.3.0. Draft iterations are tagged
ax25-v2.3.0.0, .1, .2, ... with ax25-v2.3.0 reserved for the release.

The changelog row recording Bill's original "Version 2.3 Revision 1"
re-versioning is kept as written (historical record); the renumbering
is recorded as a new row appended per the table's convention.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XodLuT1R8a4HY76RSZrqFi
@M0LTE M0LTE changed the title Rename v2.3 draft to stable filename; adopt git tags for revisions Rename v2.3 draft to stable filename; settle numbering at 2.3.0; adopt git tags Jul 16, 2026
@M0LTE
M0LTE marked this pull request as ready for review July 16, 2026 22:44
@M0LTE
M0LTE merged commit 3e8d0ca into main Jul 16, 2026
1 check passed
@M0LTE
M0LTE deleted the rename-v23-draft branch July 16, 2026 22:44
M0LTE added a commit that referenced this pull request Sep 10, 2026
* docs: findings report for the fuzzing-the-figures brief

Records what was built against docs/fuzzing-the-figures.md on 10 September 2026 and what it found. Three checkers now run in packet-net/ax25sdl CI: the static totality and determinism lint (#88), the reference interpreter with its golden traces and the clean-room SREJ traces (#75, #76), and the two-station bounded explorer with its calibration suite (#89).

Headline results: the figures are total and deterministic (8051 feasible valuations, no holes, no overlaps, also true before the #40 fix, so static totality could not have found #40); the calibration score is 10 of the 11 known defects rediscovered between the explorer's invariants and the golden traces, with only #41 out of reach; and one novel hypothesis, H1, a station entering Timer Recovery with no drawn path out once the clearing acknowledgement arrives on an I frame, corroborated by an author's note in direwolf's source. The report also carries the citation coverage map (Timer Recovery has 38 of 92 arms uncited), the pinned delegated semantics, the decisions taken against the brief's suggestions, the model's limits, and the recommended next steps. Nothing is filed as an erratum; the brief's report-first contract holds.

The brief gets a one-line status pointer to the findings.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014qyx5dT8rmRTb8szT4SpEQ

* docs(findings): H1 reproduced on packet.net; runtime acknowledge-pending defect

Folds in the packet.net two-station harness reproduction (packet-net/packet.net#811): the Timer Recovery stranding happens on the real runtime in both quirk modes, a T3 expiry is dropped because figc4.5 has no arm for it, and with the figure run as drawn the retry counter ratchets to N2 across recoveries on a working link. The control case with the acknowledgement on an RR completes normally, so only the ack carrier differs. The same trace exposed a runtime-only defect, Acknowledge Pending left set after an inline retransmission, filed as packet-net/packet.net#812 and added to the next steps.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014qyx5dT8rmRTb8szT4SpEQ

---------

Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant