docs: Add 2026-04-30 meeting notes (#699)

Author: sam3k

notes/2026/2026-04-30.md

@@ -0,0 +1,83 @@
+# 2026-04-30 ESLint TSC Meeting Notes
+
+## Transcript
+
+[`2026-04-30-transcript.md`](2026-04-30-transcript.md)
+
+## Attending
+
+- Nicholas C. Zakas (@nzakas) - TSC
+- Milos Djermanovic (@mdjermanovic) - TSC
+- Francesco Trotta (@fasttime) - TSC
+
+@nzakas moderated, and @sam3k took notes.
+
+## Topics
+
+### Follow-up from Previous Meeting
+
+**TSC Summary:** No follow-up items to review from the previous meeting.
+
+### Statuses
+
+* **@nzakas:** Health continued to keep him from contributing in any meaningful way since the last meeting.
+* **@mdjermanovic:** Fixed some ASI-related bugs in core rules and has been working on upgrading `file-entry-cache` to v11.
+* **@fasttime:** Made progress on the `--base-path` RFC and added tests for older TypeScript versions to a few repos.
+
+### Availability Next Two Weeks
+
+* **@nzakas:** Can't be sure given his health; best to assume he won't be around consistently until things resolve.
+* **@mdjermanovic:** Hopes to be able to work 1-1.5 hours each day.
+* **@fasttime:** Hopes to work 9-12 hours per week the next couple of weeks.
+
+### RFC Duty Update
+
+* This week: @mdjermanovic
+* May 4: @nzakas
+* May 11: @fasttime
+
+### [Bug: Config loading crashes when `require.cache` is unavailable](https://github.com/eslint/eslint/issues/20813)
+
+**TSC Summary:** After @fasttime's investigation, the crash turned out to be a bug in Yarn. The team discussed whether to wait for upstream fixes or land a guard in ESLint.
+
+**Discussion Points:**
+- @fasttime suggested waiting for the next Node.js 24 release to see if it fixes the problem before deciding how to proceed.
+- @nzakas felt the fix is low-risk and preferred landing it so users are unblocked rather than waiting on Yarn or Node.js.
+- A PR is already open: [PR #20812](https://github.com/eslint/eslint/pull/20812).
+
+**Resolution:** The TSC agreed to move forward with [PR #20812](https://github.com/eslint/eslint/pull/20812) so users aren't blocked while Yarn/Node.js work on their own fixes.
+
+### [fix: correctly identify file paths in `loadFormatter`](https://github.com/eslint/eslint/pull/20814)
+
+**TSC Summary:** The PR author flagged that the change could break some existing usage and asked whether it should be treated as a breaking change for v11.
+
+**Discussion Points:**
+- @nzakas: if any existing use cases stop working, the change should wait until v11.
+- @fasttime agreed that's the safer route.
+
+**Resolution:** The TSC agreed this is a breaking change and should be held until v11.
+
+### [Change Request: Checking `meta.deprecated` schema in rule tester](https://github.com/eslint/eslint/issues/20603)
+
+**TSC Summary:** The team discussed whether and how invalid `meta.deprecated` values in rules should be handled. (Noted that "tsc agenda" is for meeting discussion and "tsc waiting" is for offline resolution; this one was picked up since the team had time.)
+
+**Discussion Points:**
+- @nzakas: because `meta.deprecated` doesn't generally affect runtime behavior, runtime validation isn't appropriate; `RuleTester` is a better place for this. By contrast, `meta.languages` is validated at runtime because it does affect runtime behavior.
+- @nzakas suggested this is a good use case for the TypeScript `JSRuleDefinition` type.
+- @mdjermanovic noted that invalid values can in fact crash ESLint at runtime — specifically, if `meta.deprecated.replacedBy` exists but isn't an array, ESLint will crash (see [eslint.js#L148](https://github.com/eslint/eslint/blob/3ffb14ea517de750ed1181579ef844af342e4096/lib/eslint/eslint.js#L148)). This was previously hit by a rule in `eslint-plugin-eslint-comments` ([fix commit](https://github.com/eslint-community/eslint-plugin-eslint-comments/commit/de049807e525ac3e2d164c8873f9035b08eb23a6)).
+- @fasttime confirmed types should already catch incorrect values during development, but `RuleTester` currently has no validation.
+- The team aligned on a two-part approach: at runtime, guard against non-array `replacedBy` (so ESLint doesn't crash) as a regular fix; in `RuleTester`, add comprehensive validation of `meta.deprecated` (a breaking change).
+- On RFC necessity: @nzakas didn't feel it warranted an RFC but didn't feel strongly. @mdjermanovic was slightly in favor; @fasttime deferred to @mdjermanovic. The team agreed to do an RFC, expecting details around invalid properties, URLs, etc.
+
+**Resolution:** The TSC agreed to:
+- Add a runtime guard so ESLint doesn't crash when `meta.deprecated.replacedBy` isn't an array (regular fix).
+- Add comprehensive `meta.deprecated` validation to `RuleTester` (breaking change), driven by an RFC.
+
+### Scheduled Release for May 1st, 2026
+
+**Packages to Release:**
+- `eslint` (no pending releases for ESLint dependencies; just the `eslint` package).
+
+**Action Items:**
+- @mdjermanovic will do the release on May 1st.
+- @mdjermanovic will re-check [PR #20812](https://github.com/eslint/eslint/pull/20812) before the release; the remaining suggestion is minor (test-related). @fasttime will also take a look.