Nectenda/nectenda

Watch

Fork 0

NEC-118: The directory forbids disabling its rules — take the suppressions back out #178

Merged

nectenda-agent merged 1 commit from worktree-nec-118-take-the-forbidden-lint-suppressions-bac into main

2026-09-25 22:37:31 +01:00

nectenda-agent commented

2026-09-25 22:31:04 +01:00

Collaborator

Task: NEC-118 — https://projectron.nerchure.com/tasks/118

The branch preview scan of cebe7cd came back Failed, with three errors that
were not there before:

Disabling 'no-restricted-globals' is not allowed.
Disabling 'obsidianmd/prefer-file-manager-trash-file' is not allowed.
Disabling 'obsidianmd/no-global-this' is not allowed.

The directory refuses the suppression of its own rules. So the
eslint-disable comments NEC-117 added — to keep the local report quiet on
findings we refuse — turned three tolerated warnings into three hard errors, and
a one-error rejection into a three-error one. The refusals are unchanged and
still right; writing them as suppressions was the mistake.

Every directive is gone, every argument stays. eslint now reports eight
warnings on the published packages and exits 0 — warnings do not fail it —
and the directory reports the same eight.

The stylelint disables go too, even though stylelint disables are not
forbidden: they made our CSS report clean while the directory reported three
warnings, and the property worth protecting over a quiet report is that our gate
reproduces their scorecard. Both now say the same three.

What the same scan confirms NEC-117 got right, which is why this is a
correction and not a revert: no-static-styles-assignment and
prefer-create-el are gone from the report, and dependencies pass. The original
blocker is fixed.

Also restores an eslint-disable for @typescript-eslint/no-this-alias that had
nothing to do with any of this. The removal matched every
eslint-disable-next-line in the files it touched rather than the three rules it
meant to. Caught by arithmetic — twelve directives removed against eleven added,
and eslint then reporting one error where the count said none.

The gap this exposes, recorded in docs/releasing.md: suppression was
standing in for a baseline of accepted findings and cannot do that job here,
because the mechanism itself is refused. A real baseline has to compare against a
committed register without changing the bytes the directory scans. Follow-up, not
this card.

Next: re-sync the mirror and re-run the branch scan, before any tag exists.

Changelog

NONE

Task: NEC-118 — https://projectron.nerchure.com/tasks/118 The branch preview scan of `cebe7cd` came back **Failed**, with three errors that were not there before: ``` Disabling 'no-restricted-globals' is not allowed. Disabling 'obsidianmd/prefer-file-manager-trash-file' is not allowed. Disabling 'obsidianmd/no-global-this' is not allowed. ``` The directory refuses the **suppression** of its own rules. So the `eslint-disable` comments NEC-117 added — to keep the local report quiet on findings we refuse — turned three tolerated warnings into three hard errors, and a one-error rejection into a three-error one. The refusals are unchanged and still right; writing them as suppressions was the mistake. Every directive is gone, every argument stays. `eslint` now reports eight warnings on the published packages and **exits 0** — warnings do not fail it — and the directory reports the same eight. The stylelint disables go too, even though stylelint disables are *not* forbidden: they made our CSS report clean while the directory reported three warnings, and the property worth protecting over a quiet report is that our gate reproduces their scorecard. Both now say the same three. **What the same scan confirms NEC-117 got right**, which is why this is a correction and not a revert: `no-static-styles-assignment` and `prefer-create-el` are gone from the report, and dependencies pass. The original blocker is fixed. Also restores an `eslint-disable` for `@typescript-eslint/no-this-alias` that had nothing to do with any of this. The removal matched every `eslint-disable-next-line` in the files it touched rather than the three rules it meant to. Caught by arithmetic — twelve directives removed against eleven added, and eslint then reporting one error where the count said none. **The gap this exposes**, recorded in `docs/releasing.md`: suppression was standing in for a baseline of accepted findings and cannot do that job here, because the mechanism itself is refused. A real baseline has to compare against a committed register without changing the bytes the directory scans. Follow-up, not this card. Next: re-sync the mirror and re-run the branch scan, before any tag exists. ## Changelog NONE

nectenda-agent added the

area:plugin

label

2026-09-25 22:31:04 +01:00

nectenda-agent added 1 commit

2026-09-25 22:31:04 +01:00

Take the suppressions back out: the directory forbids disabling its rules

Release note / release-note (pull_request) Successful in 12s

Details

CI / build (pull_request) Successful in 4m38s

Details

CI / e2e (pull_request) Successful in 5m15s

Details

CI / promote (pull_request) Has been skipped

Details

CI / build (push) Successful in 4m54s

Details

CI / e2e (push) Successful in 5m16s

Details

CI / promote (push) Successful in 31s

Details

8c088366a7

The branch preview scan of `cebe7cd` came back Failed, with three errors that
were not there before:

  Disabling 'no-restricted-globals' is not allowed.
  Disabling 'obsidianmd/prefer-file-manager-trash-file' is not allowed.
  Disabling 'obsidianmd/no-global-this' is not allowed.

The `eslint-disable-next-line` comments added in the previous commit — to keep
the local report quiet on findings we refuse — are themselves rejected. So three
tolerated warnings became three hard errors, and a one-error rejection became a
three-error one. The refusals are unchanged and still right; expressing them as
suppressions was the mistake.

Every directive is gone and every argument stays. `eslint` now reports eight
warnings on the published packages and exits 0, because warnings do not fail it,
and the directory reports the same eight.

The stylelint disables go too, although stylelint disables are *not* forbidden.
They made our CSS report clean while the directory reported three warnings, and
the property worth protecting over a quiet report is that our gate reproduces
their scorecard. Both now say the same three.

Two things the same scan confirms the previous commit got right, which is why
this is a correction and not a revert: `no-static-styles-assignment` and
`prefer-create-el` are gone from the report, and dependencies pass.

Also restores an `eslint-disable` for `@typescript-eslint/no-this-alias` that
had nothing to do with any of this. The removal matched every
`eslint-disable-next-line` in the files it touched rather than the three rules
it meant to, and took a legitimate one with it — caught because eslint then
reported one error where the count said there should be none. Twelve directives
removed against eleven added is the arithmetic that gave it away.

Task: NEC-118

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01P5JXvAgQ27c4dMFDB2Wu5B

nectenda-agent merged commit 8c088366a7 into main

2026-09-25 22:37:31 +01:00

No reviewers

No milestone

No project

No assignees

1 participant

Notifications

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set

Reference

Nectenda/nectenda!178

No description provided.

Rows
Columns