Repository navigation
perf: warm lazy surfaces, lighter remote snapshots and cheaper streaming - #18979
Open
maria-rcks wants to merge 74 commits into
Open
maria-rcks wants to merge 74 commits into
maria-rcks wants to merge 74 commits into
Conversation
github-actions
Bot
added
the
vouch:trusted
juliusmarminge
added
the
macroscope-review
github-actions
Bot
added
the
size:XL
![macroscopeapp[bot]](https://avatars.githubusercontent.com/in/900172?s=60&v=4){kind=small}
# Conflicts: # apps/server/src/orchestration-v2/ProjectionStore.test.ts # apps/server/src/orchestration-v2/ThreadReadPayload.ts # apps/server/src/orchestration-v2/http.ts # apps/web/src/components/ChatMarkdown.tsx # apps/web/src/components/CommandPaletteHost.tsx # apps/web/src/components/ComposerPromptEditorTiptap.tsx # apps/web/src/components/chat/useComposerMultilinePrompt.ts # apps/web/src/markdown-incremental.test.tsx # packages/client-runtime/src/state/shell-sync.test.ts
github-actions
Bot
added
size:L
|
|
||
| const loadShellSnapshot = Effect.fn("http.orchestration.loadShellSnapshot")(function* ( | ||
| compactPullRequests: boolean, | ||
| deduplicatePullRequests: boolean, |
Contributor
There was a problem hiding this comment.
🟡 Medium orchestration-v2/http.ts:80
When deduplicatePullRequests is true, shellSnapshot still returns every thread’s pull-request list in full, so clients receive none of the requested pullRequestReference savings. The parameter is accepted by loadShellSnapshot but never used; apply the reference transformation to the snapshot before returning it.
🚀 Reply "fix it for me" or copy this AI Prompt for your agent:
In file @apps/server/src/orchestration-v2/http.ts around line 80:
When `deduplicatePullRequests` is `true`, `shellSnapshot` still returns every thread’s pull-request list in full, so clients receive none of the requested `pullRequestReference` savings. The parameter is accepted by `loadShellSnapshot` but never used; apply the reference transformation to the snapshot before returning it.
# Conflicts: # apps/server/src/orchestration-v2/ProjectionStore.ts # apps/server/src/telemetry/AnalyticsService.test.ts # apps/web/src/components/Sidebar.logic.ts # apps/web/src/components/Sidebar.tsx # apps/web/src/components/chat/V2ItemInspector.tsx # apps/web/src/terminal/ghostty/surface.ts # packages/client-runtime/src/state/shell-sync.test.ts
This branch has not been deployed
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 free
to 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.
stacked on #18220. this diff shows only this layer.
terminal and diff assets warm before use and reveal without the suspense fallback delay. plain single-line streams skip markdown tokenization, and composer key handling reuses cursor bounds. shell snapshots share repeated pull request lists, reconnects resume synchronized state, and thread switches stop showing the previous conversation under a new title.
historical medians of five linux/chromium runs before trimming:
the trim recheck kept terminal first use at 198.7 → 194.9 ms. slow-link parity remains unverified: at 5 mbit/s, editable-composer medians rose 4,883 → 5,258 ms and reconnect 3,834 → 3,923 ms, with broad overlapping ranges.
model: claude-opus-5-5 coordinating gpt-6.1-sol subagents, harness: claude code in t3 code.