From c3409a1a7f060504c2708ba023c9fe35ebe60803 Mon Sep 17 00:00:00 2001
From: "Jonathan D.A. Jewell" <6759885+hyperpolymath@users.noreply.github.com>
Date: Thu, 17 Sep 2026 19:20:36 +0000
Subject: [PATCH] chore(canon): re-adopt canon 2.0.2
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The canon's 2.0.2 release corrects the machine tree's spelling in its two
normative artefacts: `machine-readable/` → `.machine_readable/`. Ruling R-A
settled the spelling; the law had been naming the minority form.

Re-pinned, all three values:

    version          2.0.1  -> 2.0.2
    criteria_sha256  efd024ad… -> 7804d77127001ed50263f0783d8a11f7211e72cb5e9f6443f2acde4b08d9ab1e
    gates_sha256     8b1a77669c9a… -> 6a8b9b420519dccdd2c586da06a9a302bdbbb3dc917f49e4a7b308882471d68a

ORDER MATTERS, AND THIS HALF GOES FIRST

[canon].order is spine-adopts-then-canon-releases: the canon cannot land 2.0.2
while pinning hashes this repo has not adopted. Gate A fails the canon until
this is in.

So **this change is red until the canon releases** — the `Canon lockstep` job
compares all six values (the F8 fix) and canon main still says 2.0.1. The two
land together, as the Phase 1 pair did. That is the ordering rule working, not a
broken gate.

F8 is why this is safe to attempt at all: before it, `gates_sha256` was recorded
and never compared, so a half-done re-pin would have been invisible. Six values,
two directions, empty reads refused.

Witness: the spine's own extractors, run against the canon tree at the 2.0.2
release, print MATCH for all three pins.
---
 .machine_readable/rsr-profile.a2ml | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/.machine_readable/rsr-profile.a2ml b/.machine_readable/rsr-profile.a2ml
index 22c03c6f..39f60ec3 100644
--- a/.machine_readable/rsr-profile.a2ml
+++ b/.machine_readable/rsr-profile.a2ml
@@ -96,15 +96,15 @@ capabilities = [
 # build/just/repo-init.just: every minted repo records the canon it was cut
 # from, so "which repos are on canon 1.x?" becomes a grep instead of a campaign.
 # ---------------------------------------------------------------------------
-version         = "2.0.1"
+version         = "2.0.2"
 # MUST equal canon.lock [canon.artifacts].criteria.sha256
-criteria_sha256 = "efd024ad9cbdf0d36d4dbce7e491531dd4ccccbfc747a04223ce4149f4b9a53d"
+criteria_sha256 = "7804d77127001ed50263f0783d8a11f7211e72cb5e9f6443f2acde4b08d9ab1e"
 # MUST equal canon.lock [canon.artifacts].gates.sha256
 # NB this hash is the gate table WITH the role = "canon" section (canon PR
 # "feat(gates): add role=canon"). Before that section existed the canon could
 # not be scored by the checker it ships, so a spine pinning the older table
 # would be pinning a canon whose law was not subject to itself.
-gates_sha256    = "8b1a77669c9ab78e9081669dbb2ab141d8af29700affc327f3dcde99f3dcd243"
+gates_sha256    = "6a8b9b420519dccdd2c586da06a9a302bdbbb3dc917f49e4a7b308882471d68a"
 lockstep_since  = "2026-09-17"
 # The order matters and is enforced server-side: a canon change that this repo
 # cannot satisfy is a change made in the WRONG ORDER, not a rejected change.