You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Commit 5ca63b0
Browse filesBrowse the repository at this point in the historyBrowse files
Restore the row name "overlap with #15259" so the five existing
"overlap row" references in README and HANDOFF resolve again. Fix a
dangling modifier (Deno 2 removed the --lock-write flag, not lock.sh),
split a "so" that made the update-switch name look like the cause of
the task shrinkage, and note that Task 8's "Configuring without the
cache" section moves with Task 7's deno-cache input instead of having
fully landed.
Copy file name to clipboardExpand all lines: dev-docs/runtime-migration/README.md
+1-1Lines changed: 1 addition & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -145,7 +145,7 @@ The rows below are the ones that still need a named person or an external answer
145
145
| PR 15 Task 2 | The replacement `port.ts` listing has `4160, // sieve` and a private `isPortSafe`. Since #15152 the list has `4190` (the Fetch standard's bad port, which Deno 2.9's `fetch` refuses) and `isPortSafe` is exported for `tests/unit/core/port.test.ts`. Retain port `4190` and the exported `isPortSafe`, and ensure `tests/unit/core/port.test.ts` still passes. | executor | step 0 of PR 15 |
146
146
| PRs 15, 16, 23 and 31, sanitizer checks | Deno 2.9 disables test sanitizers by default, and `tests/test.ts` does not enable them. Enabling them causes every test to fail because the harness leaks an open file and a timer; these leaks also occur on Deno 2.7.14. Fix the harness leaks and re-enable sanitizers before relying on the plans' "no sanitizer failures" checks or PR 31's "Leak guarantees". | maintainer | before PR 15 runs |
147
147
| PR 00a, completion |#15186 merged on 2026-10-29; Quarto 1.11.9 shipped stable on 2026-10-30 with Deno 2.9.8. Do not execute the old bump plan again. The behavior verification is complete; #15227 fixed the cache reset caused by SQLite sidecars in the packaged seed. Building a sidecar-free seed remains a separate packaging follow-up, not an unfinished runtime bump. | coordinator | before resuming Phase 0 |
148
-
| PR 00b, reconciliation with #15259 | #15259 merged on 2026-11-03 and landed the lockfile half: committed root `deno.lock` (format v5; `tests/deno.lock` gitignored); one `deno install --frozen --allow-import` over the six entrypoints in `vendor.sh`/`vendor.cmd`, fatal on failure, with `configure.cmd` aborting on a vendor error; frozen dev launchers (`package/scripts/common/quarto`, `package/scripts/windows/quarto.cmd`), frozen `quarto-bld`/`quarto-bld.cmd` and frozen `bundle()` in `package/src/util/deno.ts`; `"lock": false` in `deno.jsonc` and `package/src/common/deno-meta.json`; `lockPath` in `denoPlugins`; `deno.lock` in the quarto-dev action cache key; `dev-docs/add-typescript-dependency.md` and `ci-deno-caching.md`. The update switch is `QUARTO_UPDATE_DENO_LOCK=1` (adopted; the plan's `update-lock.sh`/`QUARTO_DENO_LOCK_MODE=update` does not exist), so Tasks 1–3 and 8 shrink to a check of what merged. Remainder, verified against `main` on 2026-11-05: Task 4 (`deno_std` `download.sh`/`download.bat` are not frozen, `lock.sh` still passes `--lock-write`, removed in Deno 2, and `deno_std.lock` is still format v3); Task 5 (`tools/check-bundle-externals.ts` does not exist); Task 6 (`tools/check-test-imports.ts` does not exist and the four `tests/smoke/inspect` tests still import `assertObjectMatch` from `deno.land/std@0.93.0`); Task 7 (the quarto-dev action has no `deno-cache` input and `test-bundle.yml` is still the single-OS workflow: no cold cache, external-specifier grep, five renders or OS matrix); the dead `deno.land/std` `assert` import in `tools/deno-esbuild-bundle.ts` still stands (its only use is commented out). Decided 2026-11-03: 03–06 stay blocked on the full 00b gate, bundle gate included. | coordinator | PR 00b step 0 |
148
+
| PR 00b, overlap with #15259 | #15259 merged on 2026-11-03 and landed the lockfile half: committed root `deno.lock` (format v5; `tests/deno.lock` gitignored); one `deno install --frozen --allow-import` over the six entrypoints in `vendor.sh`/`vendor.cmd`, fatal on failure (`configure.cmd` propagates the error); frozen dev launchers (`package/scripts/common/quarto`, `package/scripts/windows/quarto.cmd`), frozen `quarto-bld`/`quarto-bld.cmd` and frozen `bundle()` in `package/src/util/deno.ts`; `"lock": false` in `deno.jsonc` and `package/src/common/deno-meta.json`; `lockPath` in `denoPlugins`; `deno.lock` in the quarto-dev action cache key; `dev-docs/add-typescript-dependency.md` and `ci-deno-caching.md`. The update switch is `QUARTO_UPDATE_DENO_LOCK=1` (adopted; the plan's `update-lock.sh`/`QUARTO_DENO_LOCK_MODE=update` does not exist). Tasks 1–3 shrink to a check of what merged; Task 8's two dev-docs landed in #15259's own form, and its "Configuring without the cache" section moves with Task 7's `deno-cache` input. Remainder, verified against `main` on 2026-11-05: Task 4 (`deno_std` `download.sh`/`download.bat` are not frozen, `lock.sh` still passes `--lock-write`, a flag Deno 2 removed, and `deno_std.lock` is still format v3); Task 5 (`tools/check-bundle-externals.ts` does not exist); Task 6 (`tools/check-test-imports.ts` does not exist and the four `tests/smoke/inspect` tests still import `assertObjectMatch` from `deno.land/std@0.93.0`); Task 7 (the quarto-dev action has no `deno-cache` input and `test-bundle.yml` is still the single-OS workflow: no cold cache, external-specifier grep, five renders or OS matrix); the dead `deno.land/std` `assert` import in `tools/deno-esbuild-bundle.ts` remains (its only use is commented out). Decided 2026-11-03: 03–06 stay blocked on the full 00b gate, including the bundle gate. | coordinator | PR 00b step 0 |
149
149
| PRs 09, 10 and 37, backports | 1.11 shipped before the Phase 1 codemods. Backports to `v1.11` as well as `v1.10` and older may need `--reverse`; decide from the target branch's actual call sites, not its release number. PR 09's backport checklist, PR 10's post-merge demo and PR 37's protocol must cover `v1.11`; discard the assumption that cherry-picks to 1.11 always share the migrated identifiers. | coordinator | PR 09 review, then step 0 of PRs 10 and 37 |
150
150
| PR 32 and milestone J, dependency | J is a soft prerequisite for PR 32: the parity allowlist may contain Julia tests until the subtree pull. Track J as related work, not a blocking dependency; require an empty Julia allowlist once the port is pulled. The stable-release condition for revisiting the upstream port is now met by 1.11.9; verify upstream progress before choosing the subtree to pull. | coordinator (dependency), external contact (upstream port) | PR 32 step 0 and next milestone J update |
151
151
| PR 08, output decoding | The standalone `execProcess` fix for UTF-8 characters split across stdout/stderr chunks must land before PR 08 starts. Preserve its separate streaming decoders and EOF flush, and include a split-multibyte parity case when moving the process layer. | coordinator | PR 08 step 0 |
0 commit comments