chore(deps)(deps-dev): bump @types/node from 24.13.3 to 26.1.2 - #481
chore(deps)(deps-dev): bump @types/node from 24.13.3 to 26.1.2#481dependabot[bot] wants to merge 1 commit into
Conversation
2a572fa to
6db2b2f
Compare
Bumps [@types/node](https://github.com/DefinitelyTyped/DefinitelyTyped/tree/HEAD/types/node) from 24.13.3 to 26.1.2. - [Release notes](https://github.com/DefinitelyTyped/DefinitelyTyped/releases) - [Commits](https://github.com/DefinitelyTyped/DefinitelyTyped/commits/HEAD/types/node) --- updated-dependencies: - dependency-name: "@types/node" dependency-version: 26.1.2 dependency-type: direct:development update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] <support@github.com>
6db2b2f to
8dd3a65
Compare
|
Declining this one —
Reinforced by The red checks are not the reasonThey are the usual missing Why the pin is load-bearingWith the bump applied and the lockfile synced, every gate passes:
That is not evidence the bump is safe; it is the reason the pin exists. Its job is to make newer-Node APIs unrepresentable at compile time, not to fail a test. Installing both type majors side by side and compiling the same file against each shows the difference: import { connect } from 'node:quic';
import { convertProcessSignalToExitCode } from 'node:util';
So it stops coming backThis is the fourth reopening of a settled decision — #334, #336 and #374 were closed, #340 briefly landed on 26.1.0 before |
|
OK, I won't notify you again about this release, but will get in touch when a new version is available. If you'd rather skip all updates until the next major or minor version, let me know by commenting If you change your mind, just re-open this PR and I'll resolve any conflicts on it. |
`@types/node` tracks the `engines` support floor (24) on purpose, not latest, so a newer-Node API cannot slip into the tree unnoticed — `ac552d5b` states it outright and multi-runtime.yml pins its Node leg to the same 24 for the same reason. Dependabot has no way to know that, so it re-opens the major bump every release cycle: #334, #336 and #374 were closed, #340 briefly landed on 26.1.0 before `ac552d5b` rolled it back, and #481 is the fourth reopening of a settled decision. What makes this worth a config rule rather than another close: the pin is load-bearing precisely where nothing observes it. With #481 applied locally and `bun.lock` synced, `bun run typecheck` passes and `bun test` reports 3544 pass / 0 fail across 296 files — every gate green. It cannot be otherwise, since the pin's job is to make the offending APIs unrepresentable at compile time rather than to fail a test. Compiling the same file against both type majors shows what it buys: `@types/node` 26 declares whole modules Node 24 does not have (`node:quic`, `node:ffi`) plus additions like `util.convertProcessSignalToExitCode`; under 24 that file is two compile errors, under 26 it is clean and would ship code that crashes on the runtime floor we advertise. Minor and patch updates inside the pinned major keep flowing. When the floor is raised deliberately, the pin and this rule move together, the way they did in `ac552d5b`. Closes #906
Bumps @types/node from 24.13.3 to 26.1.2.
Commits