Revert "feat(permissions): drag-to-grant onboarding for macOS TCC (Stage 1) (#1515)" - #1523
Closed
jackwener wants to merge 1 commit into
Closed
Revert "feat(permissions): drag-to-grant onboarding for macOS TCC (Stage 1) (#1515)"#1523jackwener wants to merge 1 commit into
jackwener wants to merge 1 commit into
Conversation
…age 1) (#1515)" This reverts commit 17362dc. #1515 landed red and took main with it. CI on main has failed on every commit since (`17362dcf`, then `0e80fe18` which merely inherits it), so every branch cut from main since then starts red — all three jobs, on four separate root causes, all in #1515's own files: - `format:check` — build-cursor-overlay.mjs has a copyFile call past Biome's line width. - `check-console` — permission-overlay-main.ts adds three console.warn sites that are not in the ALLOW map. - `knip` — src/overlay/permission-overlay{,-preload}.ts are unreachable from the desktop entry graph (they are bundled by the separate build:overlay esbuild step), plus an unused `asOsPermissionId` export. - `e2e` — settings.spec.ts:141 now finds no permission row rendering both grant buttons, so the window-floor contract has nothing to measure. The first two are mechanical and I had already fixed them (#1518). The knip one is a config question. But the e2e failure is a behaviour change inside #1515's own feature area — whether the fixture or the permission-row UI is wrong is a call for that change's author, not something to guess at while main stays red for everyone. Stage 1 is a new feature, not an urgent fix, so reverting is the cheap option: main goes green now and #1515 can re-land with the four checks addressed. Nothing else depends on it — the revert applied with no conflicts on top of #1359. Verified on this branch: format:check, check-console, check-a11y, check-copy, knip (apps/desktop) all clean; typecheck exit 0; `npm run test` exit 0 with 2860/2860 passing.
Member
Author
|
Closing — no longer needed. #1520 landed while this was in review and fixed main: The premise for this revert was 'main is red and #1515 is the cause'. That is no longer true, so reverting a working feature off a green main would be a net loss. #1515 stays. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Reverts
17362dcf(#1515).Why
main is red and has been since #1515 landed. This is measured on main's own CI runs, not inferred from a branch:
ade1e902(#1399)17362dcf(#1515)0e80fe18(#1359)Every branch cut from main since then starts red. Four separate root causes, all in #1515's own files:
format:check—scripts/build-cursor-overlay.mjshas acopyFile(...)call past Biome's line width.check-console—permission-overlay-main.tsadds threeconsole.warnsites not in the ALLOW map.knip—src/overlay/permission-overlay.tsandpermission-overlay-preload.tsare unreachable from the desktop entry graph (they're bundled by the separatebuild:overlayesbuild step), plus an unusedasOsPermissionIdexport. This one was hidden behind Add Rive workflow Maka tool #1 — typecheck exits atformat:checkand never reaches the knip step, so it only became visible after fixing the formatting.e2e—settings.spec.ts:141finds no permission row rendering both grant buttons, so the window-floor contract has nothing to measure.Why revert rather than fix forward
I fixed #1 and #2 mechanically in #1518, and #3 is a knip-config question I could also take. But #4 is a behaviour change inside #1515's own feature area — whether the fixture or the permission-row UI is the wrong side is a call for that change's author. Guessing at it while main stays red for everyone is the worse trade.
Stage 1 is a new feature, not an urgent fix, so reverting is cheap: main goes green now, and #1515 re-lands once the four checks are addressed. Nothing depends on it — the revert applied with no conflicts on top of #1359.
#1518 becomes unnecessary and I'll close it.
Verification on this branch