Repository navigation
chore(rtvision): sync upstream and prepare 0.0.86 - #97
Conversation
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…hanges (pingdotgg#20819) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…ngs row (pingdotgg#20830) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…ry result row (pingdotgg#20817) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…g the filter (pingdotgg#20840) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…ing (pingdotgg#20838) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…ingdotgg#20837) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
pingdotgg#20814) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…r one port (pingdotgg#20875) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…fig pushes (pingdotgg#20852) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…row (pingdotgg#20854) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…rs identical (pingdotgg#20855) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
… changed (pingdotgg#20848) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
… updates (pingdotgg#20853) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…ole page (pingdotgg#20859) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…ome (pingdotgg#20842) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
… it (pingdotgg#20891) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…otgg#20892) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
pingdotgg#20845) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
… and thread panels (pingdotgg#20823) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…dotgg#20850) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…presentation consumers (pingdotgg#20821) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
… card and model row (pingdotgg#20851) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…us returns (pingdotgg#21118) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…ingdotgg#21121) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…n unchanged preview list (pingdotgg#21130) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…nutes (pingdotgg#21122) Co-authored-by: Michał Pierzchała <thymikee@gmail.com> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…'s rows (pingdotgg#21123) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…21146) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…21147) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…alls into their threads (pingdotgg#20856) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…of its script (pingdotgg#20857) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
|
@coderabbitai review Please review exact candidate 536bcdc at the maintainer-requested high-level upstream integration scope. Review the changelog, selected conflict resolutions, retained fork behavior and obvious regressions, not exhaustive file-by-file or vendored reference content. If the 442-file diff exceeds reliable coverage, state the coverage limitation explicitly; a green check without substantive coverage is not approval. |
|
Thread transfer impact✅ Thread transfer remains within every enforced ceiling.
Baseline: Scenario and decoded snapshot size10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.
Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed. |
Final high-level primary reviewSOURCE GO for The candidate preserves merge Reviewed the changelog, selected integration/conflict resolutions, retained fork behavior, dependency changes and the targeted repair. Upstream caching, extracted components/hooks, stable timeline contexts and browser-test setup replace the older implementations. Retained fork behavior includes Gitea/Forgejo, registry/updater routing, Node/npm/OpenRC/Alpine/node-pty/SSH and packaging, route-aware editor permissions, Vim navigation, desktop profile lifetime, richer PR review/diff controls and scheduled-task permissions/model traits. Mobile CI details layer below upstream PR rows. No vendored reference content changed. CI findings and repairsInitial CI run
Scoped web/client-runtime typechecks, six-file lint, formatting and diff checks pass. Prior focused suites passed 614 tests. No repo-wide local checks or browser/computer use were performed. Independent review and CodeRabbit limitationAstra round 1 gave GO on the previous candidate, with no source findings. It does not approve the repaired SHA. Fresh final-SHA round 2 is running under task: CodeRabbit explicitly replied: "Review skipped: 440 files exceed the limit of 100." Its green check is not substantive approval. The maintainer-authorized large-diff fallback therefore applies, but requires fresh independent GPT-6 Astra GO on this exact candidate, this primary GO, all applicable CI passing and no unresolved blockers. No repeated full-review command will cure the file limit. Merge remains gated on fresh Astra GO and current CI. Release build, all expected artifacts/checksums, exact-tarball registry publication/integrity, clean Node/service-preflight and Alpine/node-pty checks must pass before GitHub draft publication. No service deployment or restart. Model: gpt-6.1-sol. Harness: Codex in T3 Code. |
High-level primary review, final fixture repairSOURCE GO for I renewed the prior review against the actual integration diff and changelog, selected timeline/settings conflicts, release manifests and dependencies. The pinned fork is Upstream settings/rendering/query improvements remain adopted. The retained Gitea/Forgejo, registry/update routing, Node/npm/OpenRC/Alpine/node-pty/SSH, release packaging, Vim navigation, editor permissions, browser profiles, richer PR controls and scheduled-task behavior are unchanged by this repair. Since the previous primary/Astra GO on The preceding CI passed lint, build/typecheck, all eight server shards, release smoke, Rust and mobile analysis. Its web job still aborted in Vitest mock initialization. The missing ESM exports are confirmed by source inspection; the new CI run must establish whether they resolve that abort. Nothing here waives a failing check. CodeRabbit explicitly skipped the large sync: "440 files exceed the limit of 100." Its green status is not approval. The authorized fallback still requires fresh exact-SHA Astra GO, this primary GO, passing applicable CI and no unresolved blockers. Fresh Astra round 3 task is No local browser/computer use, live userdata writes, service deployment or restart. Release publication remains gated on merged-head CI/build, expected installers/CLI/npm assets and checksums, registry integrity, clean Node/service preflight and Alpine/node-pty validation. Model: gpt-6.1-sol. Harness: Codex in T3 Code. |
|
Independent GPT-6 Astra round 3 returned explicit SOURCE GO for exact candidate ec14a6d. It reviewed the actual integration diff, upstream changelog, selected integration/dependency changes and all three fixture repairs since round 2. No source blockers identified. Task receipt: node:delegated-task:command%3Amcp%3Af6fb80e7-27f2-4db9-af6b-1f94753c4f5f%3Adelegate-task%3Artvision-sync-20261018-ec14a6d-astra-high-level-r3. Together with the renewed primary GO above, this supplies the two exact-SHA source approvals for the authorized large-diff CodeRabbit fallback. CodeRabbit skipped over its 100-file limit; no bot approval is inferred. CI run 39151549185 remains the next gate. Merge and all post-merge release/artifact/npm/compatibility checks remain pending. |
Primary high-level review, keybinding fixture repairSOURCE GO for exact candidate The previous web run now completes instead of aborting during mock initialization: 8,095 tests pass and two keybinding-editor fixtures fail. Every other applicable CI job passed, including shared tests, all eight server shards, lint/build/typecheck, Rust/mobile static analysis and release smoke. The two fixture failures were reproduced in a Node-only diagnostic and are not waived. The warning fixture now supplies the resolved local-exec route required by current editor dispatch. The settings fixture no longer mocks the obsolete All six tests pass both in the Node synthetic-DOM diagnostic and in the same diagnostic with React Compiler enabled. Scoped web typecheck, two-file lint/format and diff checks pass. CI still owns actual browser qualification. Refreshed pins remain fork CodeRabbit explicitly skipped the sync over its 100-file limit. Its green status is not approval. Fresh independent Astra review is running as task Model: gpt-6.1-sol. Harness: Codex in T3 Code. |
|
Independent GPT-6 Astra R4 returned explicit SOURCE GO for exact final candidate 42d7d81, with no blocking findings at the requested high-level scope. It reviewed actual integration/changelog/selected conflicts and the full two-file fixture delta. It confirmed denial coverage, permission subscriptions, actual command payload, and enabled-before-revocation controls. Task receipt: node:delegated-task:command%3Amcp%3Af6fb80e7-27f2-4db9-af6b-1f94753c4f5f%3Adelegate-task%3Artvision-sync-20261018-42d7d81-astra-high-level-r4. This and the primary GO above renew both exact-SHA source approvals for the maintainer-authorized CodeRabbit large-diff fallback. CodeRabbit skipped over its file limit, not approved. Applicable current-head CI and all post-merge release gates remain required. PR watching continues. |
|
MERGE GO for reviewed head 42d7d81. Final readiness rechecked: correct RTVision fork and rtvision base 019d696, unchanged pinned upstream 3ff5b3c, clean/mergeable PR, no review threads or unresolved blocking findings. Primary and independent GPT-6 Astra R4 source GO name this exact head. CodeRabbit explicitly skipped the 440-file sync over its 100-file limit, so the maintainer-authorized high-level primary+Astra fallback applies. Exact-head CI39152521699 and applicable checks passed; transfer limits pass for both Codex and Claude. Merge will preserve upstream ancestry. Release publication is not yet approved by build evidence: merged-head CI/release build and all artifact/npm/Node/service/Alpine gates must still pass. |
Released 0.0.86RTVision T3 Code 0.0.86 is published.
The release notes summarize the 62 upstream commits, upstream replacements, retained fork behavior and validation boundaries. Full receipts remain under Model: gpt-6.1-sol. Harness: Codex in T3 Code. |
Sync 62 upstream commits through
3ff5b3c82a62baddc0c2b1b227e9d3a06190c426intortvisionand prepare patch release 0.0.86. Fork base is019d696159b75e0ad0abe45be60937c0959dc1a7; previously integrated upstream ise6ced718a3bb35c58986b12fb9f84624974104c2. Both parent histories are preserved.Upstream changes
t3 doctorcan remove the old database explicitly while preserving the current one.Fork integration decisions
Use upstream caching, extracted components/hooks, stable timeline row contexts, and browser-test setup rather than retaining superseded fork implementations. Layer mobile CI-run details below upstream's richer linked-PR rows.
Keep RTVision Gitea/Forgejo integrations, registry and updater routing, Node/npm/OpenRC/Alpine/node-pty support, SSH/editor routing, release packaging, Vim navigation, browser profiles, richer PR review behavior and diff scroll retention, and scheduled-task model traits and permissions. The remaining fork-specific node-renderer tests still need their test-only dependencies/helpers; obsolete jsdom use is removed. No vendored reference content changed.
Gates
Model: gpt-6.1-sol. Harness: Codex in T3 Code.