Uh oh!
There was an error while loading. Please reload this page.
[world-vercel] Read hasMore in v4 list pagination - #2486
Merged
Conversation
The v4 LIST frame stream ends with a sentinel that carries a trailing `next` cursor even on the final page (it doubles as the incremental-load resume point). Deriving `hasMore = Boolean(next)` therefore reports "more pages" on every non-empty page, so the runtime's `while (hasMore)` replay loader issues one extra empty-page request to the backend on every event-log load. Read the server's explicit `hasMore` from the sentinel and use it, falling back to `Boolean(next)` for older servers that don't emit it. Pairs with the backend change adding `hasMore` to the sentinel. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Contributor
🦋 Changeset detectedLatest commit: 255caea The changes in this PR will be included in the next version bump. This PR includes changesets to release 17 packages
Not sure what this means? Click here to learn what changesets are. Click here if you're a maintainer who wants to add another changeset to this PR |
Contributor
🧪 E2E Test Results✅ All tests passed Summary
Details by Category✅ ▲ Vercel Production
✅ 💻 Local Development
✅ 📦 Local Production
✅ 🐘 Local Postgres
✅ 🪟 Windows
✅ 📋 Other
|
Contributor
📊 Benchmark Results
workflow with no steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Next.js (Turbopack) workflow with 1 step💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Next.js (Turbopack) workflow with 10 sequential steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Next.js (Turbopack) workflow with 25 sequential steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Next.js (Turbopack) workflow with 50 sequential steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Next.js (Turbopack) | Express Promise.all with 10 concurrent steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Next.js (Turbopack) Promise.all with 25 concurrent steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Next.js (Turbopack) Promise.all with 50 concurrent steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Next.js (Turbopack) | Express Promise.race with 10 concurrent steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Next.js (Turbopack) | Express Promise.race with 25 concurrent steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Next.js (Turbopack) | Express Promise.race with 50 concurrent steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Next.js (Turbopack) workflow with 10 sequential data payload steps (10KB)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Next.js (Turbopack) workflow with 25 sequential data payload steps (10KB)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Next.js (Turbopack) workflow with 50 sequential data payload steps (10KB)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Next.js (Turbopack) workflow with 10 concurrent data payload steps (10KB)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Next.js (Turbopack) workflow with 25 concurrent data payload steps (10KB)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Next.js (Turbopack) workflow with 50 concurrent data payload steps (10KB)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Next.js (Turbopack) Stream Benchmarks(includes TTFB metrics)workflow with stream💻 Local Development
▲ Production (Vercel)
🔍 Observability: Next.js (Turbopack) | Express stream pipeline with 5 transform steps (1MB)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Next.js (Turbopack) 10 parallel streams (1MB each)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Next.js (Turbopack) fan-out fan-in 10 streams (1MB each)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Next.js (Turbopack) | Express SummaryFastest Framework by WorldWinner determined by most benchmark wins
Fastest World by FrameworkWinner determined by most benchmark wins
Column Definitions
Worlds:
❌ Some benchmark jobs failed:
Check the workflow run for details. |
…t-hasmore # Conflicts: # packages/world-vercel/src/events.test.ts
pranaygp
left a comment
Contributor
There was a problem hiding this comment.
Reviewed against the merged server PR (vercel/workflow-server#539). The change is correct and well-scoped:
- Logic is right.
typeof result.hasMore === 'boolean' ? result.hasMore : Boolean(result.next)prefers the explicit flag and thetypeofguard correctly propagatesfalse(a truthiness check would have dropped it). The conditional spread...(hasMore !== undefined ? { hasMore } : {})likewise preserves the legacyundefinedsemantics the test asserts. - Cursor is preserved.
cursor: result.next ?? nullis unchanged, so incremental-load resume still works whenhasMoreisfalse. I traced both consumers:loadWorkflowRunEvents(while (hasMore)) exits one iteration earlier on the common single-page case, and sinceassertEventPaginationProgressonly runs whenhasMoreis true, emittingfalsecleanly skips the cursor-progress assertion rather than tripping it.wakeUpRun(runs.ts:154) benefits identically. - Both skew directions hold. New SDK + old server falls back to
Boolean(next); old SDK + new server ignores the additive field. Matches the server side, which is additive on both v4 list handlers. - Decode path is genuinely exercised.
decodeFramesdoes a full CBORdecode, sohasMoresurvives, and the tests round-trip through realencodeFrame/decodeFramesrather than stubbing the result.
Verified locally: the two touched test files pass (24 tests) and @workflow/world-vercel typecheck is clean. Changeset is a correctly-scoped patch.
One maintainability nit inline, plus an optional test-coverage note below. Nothing blocking.
Optional: the new getWorkflowRunEvents hasMore mapping block covers false+next, true+next, and legacy-with-next→true. It doesn't cover the legacy-server-no-next case ({_end: 1} → hasMore: false via Boolean(undefined)). It's trivially correct, but adding it would lock in the full fallback truth table.
| /** Pagination cursor — present when more pages remain. */ | ||
| /** | ||
| * Trailing cursor. Present even on the final page — it doubles as the | ||
| * resume point for incremental loads — so it is NOT a reliable "more |
Contributor
There was a problem hiding this comment.
Doc drift: the sentinel wire-shape is described in three other places that still read {_end: 1, next?: ...} and don't mention hasMore:
frames.ts:8— the ABNF wire spec (end-frame := ... cbor_meta {_end: 1, next?: string} || u32_be(0))events-v4.ts:14—sentinel frame (meta =+ "{_end: 1, next?: cursor})"events.ts:25
The server's matching frames.ts ABNF was explicitly updated to {_end: 1, next?: string, hasMore?: boolean} in vercel/workflow-server#539. Worth syncing at least frames.ts:8 here so the two protocol specs don't diverge for the next reader.
Contributor
There was a problem hiding this comment.
Fixed in 255caea — synced all three (frames.ts:8 ABNF, events-v4.ts:14, events.ts:25) to include hasMore?: boolean.
The v4 LIST end-frame now carries an explicit `hasMore` flag, but three
doc comments still described the sentinel as `{_end: 1, next?: ...}`.
Update the ABNF wire spec and the two prose mentions so they match the
server's frames.ts (vercel/workflow-server#539).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>pranaygp
approved these changes
Jun 17, 2026
pranaygp
enabled auto-merge (squash)
June 17, 2026 23:53
VaguelySerious
disabled auto-merge
June 18, 2026 00:15
Uh oh!
There was an error while loading. Please reload this page.
Contributor
No backport to This fixes the To override, re-run the Backport to stable workflow manually via |
pranaygp added a commit
that referenced
this pull request
Jun 18, 2026
…-hydration * origin/main: perf(core): lazy inline step start (save one world round-trip per step) (#2478) perf(core): skip per-step events.list via inline event-log delta (#2475) Version Packages (beta) (#2491) [world-vercel] Honor hasMore flag from v4 list pagination endpoint (#2486) Version Packages (beta) (#2451) Fix Next workflow module specifier root (#2455) [world-vercel] Send remoteRefBehavior=lazy on v4 metadata-only event listings (#2415) [swc-plugin] Fix eager discovery for object property steps (#2484) fix(web-shared): align attributes panel styling (#2483) [web-shared] Auto-scroll trace viewer on J/K span navigation (#2366) fix(web): render restarted step segment as solid gray, not running stripes (#2480) fix(web-shared): use solid gray for queued trace segment (#2474) Add trace viewer span markers for hooks and attributes (#2452) test: support Vercel protection bypass secret in e2e headers (#2458) fix(core): bump payload-compression cutoff to 5.0.0-beta.18 (#2470) Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> # Conflicts: # packages/core/src/runtime.ts
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Context
When the Vercel backend's event endpoints moved to the v4 wire format (5.0.0-beta.17), the LIST response began ending with a sentinel frame
{_end: 1, next?: cursor}. The backend returns a trailingnextcursor on the final page too — it doubles as the resume point for incremental loads — sonextis present even when no more pages remain.The SDK derived
hasMore = Boolean(next), so every non-empty page reported "more pages". The runtime's replay event loader is awhile (hasMore)loop, so it issued one extra request that returns an empty page on every event-log load — i.e. on every replay of every run. A single-page run that needed 1 list request now needs 2.Change
Read the backend's explicit
hasMorefrom the sentinel and use it, falling back toBoolean(next)for older backends that don't emit the flag. Thecursormapping is unchanged — incremental loads still get the trailing cursor.This pairs with the backend change that adds
hasMoreto the sentinel. Both sides are backward-compatible: a new SDK against an old backend keeps the (correct, if slower)Boolean(next)behavior, and an old SDK ignores the new field.Tests
events-v4.test.ts—consumeListFrameStream/getWorkflowRunEventsV4captures an explicithasMore: falseeven when a trailingnextis present; leaveshasMoreundefined for a legacy sentinel.events.test.ts—getWorkflowRunEventsmaps explicithasMoretrue/false through, keeps the cursor, and falls back toBoolean(next)against a legacy server.All
@workflow/world-vercelunit tests pass; typecheck and lint clean.🤖 Generated with Claude Code