From 1bbdde6c14ded0293f666af243acd95dfc517799 Mon Sep 17 00:00:00 2001 From: Wessel Verheij Date: Mon, 14 Sep 2026 02:23:02 +0200 Subject: [PATCH] docs(ops): the goal the note above it said was not one MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The v2.0 manifest's prose said F1-R18 "is NOT a goal of v2.0". It is: it sits in the goals list forty lines below, under the desktop shell's own heading. The note was written before the goal was added and never re-read against the list it introduces. manifest_check reads the list, not the prose, so this never showed as a gate failure — it showed as a page that tells a reader the opposite of what the machine acts on. The note now says which one is authoritative and why it was wrong, rather than being quietly deleted. Found while diffing the goal SET against the commit that audited it: the audit of 2026-09-05 covered 339 identifiers and the manifest now carries 373, so 34 goals have never been audited against the code. Eleven of those are named nowhere in the audit commentary at all. Co-Authored-By: Claude Opus 5 (1M context) Signed-off-by: Wessel Verheij --- 70-operations/versions/2.0.0.toml | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/70-operations/versions/2.0.0.toml b/70-operations/versions/2.0.0.toml index 691c0ee..a9e225c 100644 --- a/70-operations/versions/2.0.0.toml +++ b/70-operations/versions/2.0.0.toml @@ -454,9 +454,13 @@ # built on 2026-09-05 in beatrax#406 and #407, hours after this note and both # feature pages were last read. They are goals now, and the markers are cleared. # -# F1-R18 (lock on window close) was the one carry-over release-readiness item -# that is NOT a goal of v2.0. It is satisfied as of 2026-09-05, and the claim -# that it "cannot be verified outside a real bundle build" was wrong: the suite +# F1-R18 (lock on window close) is the one carry-over release-readiness item, +# and it IS a goal of this release — it is in the list below, under the desktop +# shell's own heading. This note said the opposite until 2026-09-14, having been +# written before the goal was added and never re-read against the list it sits +# above; the list is what scripts/manifest_check.py reads, so the prose was the +# half that was wrong. It is satisfied as of 2026-09-05, and the claim that it +# "cannot be verified outside a real bundle build" was wrong too: the suite # persists the window's session, drives the real shell event POST and reloads # that session from the handler. F1-R15 took its place as F1's open item — the # watchdog's rolling counter does not survive the request that writes it — and