[pull] main from pingdotgg:main - #220
Merged
Merged
Conversation
Defer the draft walk and JSON serialization until the storage write flushes. Preserve hydration, migrations, attachment verification, and final flushes. Continues the web portion of [#9049](#9049). The mobile storage migration remains separate. Created with GPT-6 Astra (preview) in Codex. Co-authored-by: Wout Stiens <71498452+StiensWout@users.noreply.github.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Append terminal history incrementally and materialize text for snapshots and coalesced disk writes. Clear evicted line references without changing retained output. Continues [#9357](#9357). The original contribution and author credit are preserved. The current line limit and wire format stay unchanged. A strict byte limit remains separate work. Created with GPT-6 Astra (preview) in Codex. Co-authored-by: will <will@moondiner.com>
Keep bounded, byte-counted terminal chunks and append only unread output. Use UTF-16 cursors so compaction preserves live terminal replies. Reset on lifecycle changes or a real retained-data gap. Keep the existing wire protocol, native buffer interface, and client retention limit. Native streaming and strict server replay byte bounds remain separate. Continue the client helpers from #9027 at source head 9391da2. Created with GPT-6 Astra (preview) in Codex. Co-Authored-By: Wout Stiens <71498452+StiensWout@users.noreply.github.com> Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Co-authored-by: maria-rcks <254055478+maria-rcks@users.noreply.github.com>
Stop post-construction terminal snapshots, canvas paint, and cursor timers while a drawer or right panel is hidden. Keep parsing output and answering VT queries, then render current state once on reveal. Cover delayed WASM initialization, selection behavior, zero-size mounts, and reveal with real-core headless tests. Keep the existing startup background fill. Continue the zero-size guard from source commit eb5b681 in the terminal streaming contribution. Created with GPT-6 Astra (preview) in Codex. Co-Authored-By: Wout Stiens <71498452+StiensWout@users.noreply.github.com> Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Accept file-backed image drafts and v4 outbox records while keeping current inline image creation and v3 outbox writes. Retain image files during previews, uploads, and legacy inline reads. Preserve image MIME types and stop canceled sends after asynchronous reads. The later file-backed writers remain held until new native runtime fingerprints contain these readers and the storage guards. Continue Wout Stiens' mobile draft work with separate reader-only corrections. The original contribution remains unchanged. Created with GPT-6 Astra (preview) in Codex. Co-Authored-By: Wout Stiens <71498452+StiensWout@users.noreply.github.com> Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Read only the selected thread's events when resuming its detail stream. Measure bounded row counts and serialized payload bytes before replay, using the existing aggregate index and a captured authoritative head. Keep snapshot resets for oversized or invalid cursors. Reset recreated threads when a snapshot exists, and keep bounded replay for deleted threads whose snapshot is absent. Shell replay and the detail-event filter are unchanged. Created with GPT-6 Astra (preview) in Codex.
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
Reuse ordered timeline entries and raw rows for safe streaming text updates. Keep full derivation for structural, metadata, activity, and control changes. Preserve immutable attachment-preview objects when their current URLs are unchanged. Cover URL renewal/removal, completion, grouping, pagination, and earlier-row immutability. Continue extoci's ordered timeline projection work with separate integration and attachment corrections. The original contribution remains unchanged. Created with GPT-6 Astra (preview) in Codex. Co-Authored-By: extoci <hi@extoci.lol>
Give HTTP sends and response-body reads ten seconds each. Keep the existing typed transport errors, delivery records, and queue policy. Abort stalled requests and let the next signed job run. JWT retrieval is outside these deadlines. Do not add retries for an uncertain response result. Created with GPT-6 Astra (preview) in Codex.
Do not retain OpenCode tool parts after emitting their runtime events. The remaining cache readers need text, reasoning, or step-usage parts, not tool input and output. Keep tool lifecycle events, output bytes, late-role assistant text, and usage handling unchanged. Add focused lifecycle coverage and verify retained memory through the real adapter. Created with GPT-6 Astra (preview) in Codex.
Co-authored-by: maria-rcks <254055478+maria-rcks@users.noreply.github.com>
Bound cached parsed review sections by eight entries and 4,194,304 full source characters. Evicted inactive parsed and native results can be collected. Raw patches, comments, and view state stay unchanged. Prewarm only nearby sections that fit beside the selected section. Use actual retained source weights for normalized cache hits, and skip oversized prewarms. Created with GPT-6 Astra (preview) in Codex.
Build one ordered terminal metadata index per immutable snapshot. Reuse unchanged session wrappers and thread groups across consumers, with separate environment targets. Keep numeric ordering, picker ties, attach state, and subscription lifetimes unchanged. Let retired snapshots and groups be collected. Created with GPT-6 Astra (preview) in Codex.
Keep at most 5,000 lines and 8 MiB of retained UTF-8 terminal history. Discard the oldest text at either limit while preserving complete live output. Track bytes and newlines in small chunks. Join split surrogates before eviction. Restore only the needed file tail, handle short reads, and close the file before rewriting current or legacy history. Created with GPT-6 Astra (preview) in Codex. Co-authored-by: will <will@moondiner.com>
Co-authored-by: maria-rcks <254055478+maria-rcks@users.noreply.github.com>
Close a thread's live detail stream when its last consumer leaves. Retain a registry-local completed state and replay cursor for five idle minutes. Keep warm lookup independent of collectible raw atom definitions. Preserve paging and deletion state, reject stale owners, and cache only completed data/cursor updates. Skip repeated saves of unchanged data while retaining dirty-data flushes and failed-write retries. Verify real GC/remount behavior and focused lifecycle cases. Created with GPT-6 Astra (preview) in Codex.
Co-authored-by: maria-rcks <254055478+maria-rcks@users.noreply.github.com>
github-actions
Bot
added
vouch:trusted
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 subscribe to this conversation on GitHub.
Already have an account?
Sign in.
8 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.
See Commits and Changes for more details.
Created by
pull[bot] (v2.0.0-alpha.4)
Can you help keep this open source service alive? 💖 Please sponsor : )