docs: bring the release runbook and the syntax reference up to 2.7.4

[PathGao]

Four claims in the docs describe behaviour that PRs since 2.7.3 changed, two features shipped without reaching the file whose job is to say what Markpad can do, and the release runbook had drifted into being something else. npm test 904 pass, npm run test:vitest 365 pass.

1. Claims that were no longer true

file	said	is
RELEASING.md	snapcraft.yaml uses npm ci and builds the snap outside Actions	it unpacks the release's own .deb (#579) and compiles nothing (#577)
RELEASING.md	snap job fails ⇒ the rust-deps part is gone or renamed	there is no rust-deps part and no rustup; the live failures are a missing markpad.deb and the snapcraft snap moving under us
RELEASING.md, README.md, README.zh-CN.md	Check for Updates… offers .deb/.rpm users an update and then fails to install it	#573 asks self_update_supported before the check and says where updates come from
README.zh-CN.md	.deb and .rpm users update "通过各自的分发渠道"	there is no apt or dnf repository — this is the sentence #566 removed from the English README, in the file it did not touch
snapcraft.yaml	.github/workflows/test_snap.yml puts the .deb there	that workflow was deleted in #601

2. The runbook is a runbook again

RELEASING.md is what you read on the day you cut a release. Over three days in August I threaded post-mortems through it: which run pushed Chocolatey 2.7.2 at 15:37 UTC, that the Snap Store served 2.6.11 for three months and six versions, which two of the three v2.7.2 attempts the AppImage strip failed in, how the 2.7.2/2.7.3 lock skew was found. All true, none of it changes what you do next, and it sits between the reader and step 5 while a build is running.

The test applied to every sentence: can it change an action taken on release day.

• Gone — the "Why package managers publish after the release" section entirely; its one operational sentence (a workflow_dispatch against a published tag is the only way to exercise snapcraft.yaml) moves into step 7, where you would reach for it. The failure histories in three troubleshooting rows, keeping the symptom and the fix. The placeholder-pubkey era. Two design defences.
• Kept — reasons that stop you doing the wrong thing: why nothing may be created at alecdotdev/Markpad, why the Cargo.lock bump is the one that gets forgotten. A prohibition without its reason gets deleted as superstition by whoever comes next.

1957 → 1563 words. The version before I started on it was 883.

The same edit removed the last issue number from both READMEs. #570 was a closed bug report, and after #573 it hung off a sentence describing ordinary behaviour — a user asking "how does my .deb update" clicked through to a fault that no longer exists. The Report a Bug link, which is what both files had before, stays.

3. The syntax reference

samples/markdown-syntax.md is what the README points users and their AI assistants at for "what can Markpad do". Two features landed since it was written:

• Lists — Enter continues the marker, Tab/Shift+Tab change level, Enter on an empty item leaves the list (feat(editor): continue a Markdown list when Enter is pressed inside one #636); the three list chords (feat(shortcuts): Mod+K inserts a link, and the table verbs move off the chord #652).
• Tables — Tab/Shift+Tab between cells, Mod+Enter rows, Mod+Shift+C / Mod+Shift+Backspace columns, Mod+Alt+T to insert one (feat(editor): add and remove table rows and columns without editing pipes by hand #645, feat(shortcuts): keys for the two column verbs, and a better rule for why #653).
• Links — Mod+K (feat(shortcuts): Mod+K inserts a link, and the table verbs move off the chord #652).

Every key was read out of Editor.svelte rather than out of a PR description — #645's Mod+K T and Mod+K C were superseded by #653 two PRs later, so the descriptions would have documented a keymap that was two days stale.

That file carries no issue numbers, no history and no defence of a decision in its 500 lines; where it does give a reason — the three deliberate incompatibilities — the reason tells the reader what to write instead. Two sentences of mine argued the design at the reader (that four other editors use Mod+K; why deleting a row has no chord while deleting a column does) and came back out. Both language versions carry the same additions.

🤖 Generated with Claude Code