Skip to content

Commit 098a9b7

Browse files
opencodeclaude
andcommitted
fix(app): split the KNOWN_MISSING pin by kind, not one combined total
Henry's review on #40: pending-first-attempt now dominates (83% of the list) and spikes with every feature that adds locale keys, so a future PR quietly adding a handful of verified-failed entries inside a thousand-entry pending bump would be invisible against a single total -- and that handful is exactly the signal this mechanism exists to catch. verified-failed stays near-static; pending-first-attempt is expected to breathe with feature work. Two numbers, each bumped deliberately for its own reason. 319 verified-failed (unchanged), 1525 pending-first-attempt (was 1844 combined). Co-Authored-By: Claude Sonnet 5 <[email protected]> Claude-Session: https://claude.ai/code/session_01JJQrqYMv43CrdbjpTT7884
1 parent 51216ff commit 098a9b7

1 file changed

Lines changed: 21 additions & 11 deletions

File tree

‎packages/app/src/i18n/parity.test.ts‎

Lines changed: 21 additions & 11 deletions
Original file line numberDiff line numberDiff line change
@@ -2001,22 +2001,32 @@ const KNOWN_MISSING: readonly (
20012001

20022002
describe("i18n parity", () => {
20032003
/**
2004-
* Pinned total, same pattern as `EventManifest.Latest.size` (packages/schema/test/
2004+
* Pinned PER KIND, same pattern as `EventManifest.Latest.size` (packages/schema/test/
20052005
* event-manifest.test.ts) -- nothing else watches this list's size, so without a pin it grows
20062006
* by ambient drift: every future feature adds its own individually-justified entries while
20072007
* overall parity quietly degrades with no signal anywhere. Growing the list is now a deliberate,
2008-
* same-commit bump of this number (lead ruling, TKT-323 chunk 3, 2026-08-12) -- the loosening
2009-
* this pin guards is this same PR's own creation (319 verified-failed + 183 pending-first-
2010-
* attempt = 502). Per-locale pins are deliberately NOT required: TKT-373's real translation
2011-
* pipeline run will collapse this total massively, and the pin update landing in that same
2012-
* commit is the mechanism working as designed, not a violation of it.
2008+
* same-commit bump of these numbers (lead ruling, TKT-323 chunk 3, 2026-08-12).
20132009
*
2014-
* Bumped 502 -> 1844 (+1342 = 22 new keys x 61 app locales) for TKT-323's MCP config editing
2015-
* add/edit/remove UI, same-commit, same TKT-373 folding, no new reasoning needed -- the pin
2016-
* doing exactly what it's for.
2010+
* Split by kind (Henry's review, PR #40) rather than one combined total: `pending-first-attempt`
2011+
* dominates (83% of the list) and spikes with every feature that adds locale keys, so a future
2012+
* PR quietly adding a handful of `verified-failed` entries inside a thousand-entry pending bump
2013+
* would be invisible against a single total -- and a handful of newly-untranslatable strings is
2014+
* exactly the signal this mechanism exists to catch. `verified-failed` should stay near-static
2015+
* (a real translation attempt actually ran and failed); `pending-first-attempt` is expected to
2016+
* breathe with feature work and collapse when TKT-373's real translation pipeline runs. Two
2017+
* numbers, each bumped deliberately for its own reason, not one number hiding two different
2018+
* kinds of drift.
2019+
*
2020+
* verified-failed: 319 (unchanged by this PR). pending-first-attempt: 183 -> 1525 (+1342 = 22
2021+
* new keys x 61 app locales) for TKT-323's MCP config editing add/edit/remove UI, same-commit,
2022+
* same TKT-373 folding, no new reasoning needed -- the pin doing exactly what it's for.
20172023
*/
2018-
test("KNOWN_MISSING's total size is a deliberate, same-commit bump, never ambient drift", () => {
2019-
expect(KNOWN_MISSING.length).toBe(1844)
2024+
test("KNOWN_MISSING's verified-failed count is a deliberate, same-commit bump, never ambient drift", () => {
2025+
expect(KNOWN_MISSING.filter((entry) => entry.kind === "verified-failed").length).toBe(319)
2026+
})
2027+
2028+
test("KNOWN_MISSING's pending-first-attempt count is a deliberate, same-commit bump, never ambient drift", () => {
2029+
expect(KNOWN_MISSING.filter((entry) => entry.kind === "pending-first-attempt").length).toBe(1525)
20202030
})
20212031

20222032
test("non-English locales have every English key and required plural variants", async () => {

0 commit comments

Comments
 (0)