Release v10.0.0-rc.19 — version bump#1245
Conversation
10.0.0-rc.18 → 10.0.0-rc.19 across all 19 workspace package.json (16 public validated == rc.19 for the npm-publish gate). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
| { | ||
| "name": "@origintrail-official/dkg-random-sampling", | ||
| "version": "10.0.0-rc.18", | ||
| "version": "10.0.0-rc.19", |
There was a problem hiding this comment.
🔴 Bug: The manifest is bumped to 10.0.0-rc.19, but this package still exports RANDOM_SAMPLING_PACKAGE_VERSION = '10.0.0-rc.2' from packages/random-sampling/src/index.ts, and its test asserts that stale value. That means the current validation would pass while the published API reports the wrong release version; update the constant or derive it from package metadata, and make the test compare against the manifest rather than a hard-coded old RC.
| { | ||
| "name": "@origintrail-official/dkg-adapter-openclaw", | ||
| "version": "10.0.0-rc.18", | ||
| "version": "10.0.0-rc.19", |
There was a problem hiding this comment.
🔴 Bug: This published adapter package includes openclaw.plugin.json in its files list, but that shipped plugin manifest still says "version": "10.0.0-rc.2". Consumers or OpenClaw registry/update logic reading the plugin manifest will see the adapter as rc.2 even though npm installs rc.19, and there is no validation covering this secondary version field.
…ifest bump The version-bump only updated package.json manifests; these three version strings live outside package.json and were left at rc.2 (flagged in #1245 review): - packages/random-sampling/src/index.ts RANDOM_SAMPLING_PACKAGE_VERSION - packages/random-sampling/test/skeleton.test.ts (the assertion on it) - packages/adapter-openclaw/openclaw.plugin.json (shipped in the package `files` list, so consumers/registry would otherwise read the adapter as rc.2) Deliberately NOT changed: GOSSIP_ENVELOPE_VERSION ('10.0.0' — a wire-protocol version decoupled from the rc release cadence) and the rc.10/rc.11 references in code comments (historical facts). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…ifest bump The version-bump only updated package.json manifests; these three version strings live outside package.json and were left at rc.2 (flagged in #1245 review): - packages/random-sampling/src/index.ts RANDOM_SAMPLING_PACKAGE_VERSION - packages/random-sampling/test/skeleton.test.ts (the assertion on it) - packages/adapter-openclaw/openclaw.plugin.json (shipped in the package `files` list, so consumers/registry would otherwise read the adapter as rc.2) Deliberately NOT changed: GOSSIP_ENVELOPE_VERSION ('10.0.0' — a wire-protocol version decoupled from the rc release cadence) and the rc.10/rc.11 references in code comments (historical facts). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
|
Addressed the rc.2 version-string feedback in
Deliberately left unchanged (flagging so it's clear it wasn't an oversight):
|
| */ | ||
|
|
||
| export const RANDOM_SAMPLING_PACKAGE_VERSION = '10.0.0-rc.2'; | ||
| export const RANDOM_SAMPLING_PACKAGE_VERSION = '10.0.0-rc.19'; |
There was a problem hiding this comment.
🟡 Issue: RANDOM_SAMPLING_PACKAGE_VERSION manually duplicates the package version, and the updated test repeats the same hard-coded literal instead of checking packages/random-sampling/package.json. This creates another release-bump hotspot that can drift from the published metadata; the previous value was already stale relative to the package version. Consider deriving or generating this value from package metadata, or at least make the test compare against package.json so future bumps fail when this export is missed.
| describe('@origintrail-official/dkg-random-sampling — skeleton', () => { | ||
| it('exposes a version constant matching the rest of the workspace', () => { | ||
| expect(RANDOM_SAMPLING_PACKAGE_VERSION).toBe('10.0.0-rc.2'); | ||
| expect(RANDOM_SAMPLING_PACKAGE_VERSION).toBe('10.0.0-rc.19'); |
There was a problem hiding this comment.
🟡 Issue: This test claims the exported version matches the rest of the workspace, but it only compares the source constant to another hard-coded 10.0.0-rc.19 literal. If package.json or another workspace package drifts, the test still passes as long as this assertion is updated with the source constant, so it does not provide validation evidence for the release-version consistency this PR is changing. Consider deriving the expected value from package/workspace metadata instead of duplicating the target string.
rc19 release — version bump
Version-bump PR for v10.0.0-rc.19. Cuts from current
main(525004a04). Single commit:10.0.0-rc.18 → 10.0.0-rc.19across all 19 workspacepackage.json(16 public validated == rc.19 for the npm-publish gate). No code or lockfile changes.What rc19 contains (already merged to
mainsince rc18)Anti-spam / security
Scalability
Observability
/api/status+ reusable load probe (feat(daemon): surface admission-control stats on /api/status + load probe #1230)Mainnet prep & stability
POST /api/assertion/:name/writereturns 500 + TypeError when given N-Quad strings instead of quad objects #306 (fix: three pre-mainnet blockers — write-route 500 crashes (#306/#787) + public-CG host-mode ingest (#1124) #1239); P0 / high pre-mainnet issues (fix: 5 P0 pre-mainnet issues (#11, #1078, #1121, #936, #1098) #1228, fix: high/pre-mainnet issues — query scoping (#184/#675), curator gate (#757), skill ACL (#462), async honesty (#1013) #1132)After merge (release process — for reference, not part of this PR)
Tag
v10.0.0-rc.19→npm-publish(reviewer-gated) → GitHub release → update testnet nodes + validate.🤖 Generated with Claude Code