From 73fbbd392290e2f2f92614ff7d118879b4c30c90 Mon Sep 17 00:00:00 2001
From: Daniel Samson <12231216+daniel-samson@users.noreply.github.com>
Date: Sun, 9 Aug 2026 17:27:43 +0100
Subject: [PATCH] =?UTF-8?q?docs:=20record=20the=20V2=20sequencing=20?=
 =?UTF-8?q?=E2=80=94=20medium=5Fchanged=20emitted=20in=20V2b,=20tested=20i?=
 =?UTF-8?q?n=20V4?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 docs/volume-manager-plan.md | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/docs/volume-manager-plan.md b/docs/volume-manager-plan.md
index edecb60..92dd5e4 100644
--- a/docs/volume-manager-plan.md
+++ b/docs/volume-manager-plan.md
@@ -49,6 +49,14 @@ harness once, not fat and the harness both.
 - Discrimination: a fixture transferring outside its assigned range must be
   refused; fails against a driver without the clamp.
 
+**Sequencing (as executed).** V2a landed the range clamp + its discrimination
+fixture (block-range) — the security mechanism is testable in isolation. V2b
+adds `medium_changed` and makes usb-storage a publisher that emits it on
+presence transitions, but its END-TO-END test (eject → medium_changed →
+unmount/remount) lands in V4 with the real consumer, the volume manager —
+rather than a throwaway subscriber fixture V3 would immediately replace. Same
+work, no duplicated scaffolding.
+
 ## V3 — the volume manager service
 
 New binary `system/services/volume-manager`, spawned by init, serving the