fix: harden desktop recording and rendering for v0.6 - #2172
Open
richiemcilroy wants to merge 24 commits into
Open
fix: harden desktop recording and rendering for v0.6#2172richiemcilroy wants to merge 24 commits into
richiemcilroy wants to merge 24 commits into
Conversation
richiemcilroy
marked this pull request as ready for review
August 30, 2026 12:35
Manage your Superagent protectionSuperagent has paused scans for this repository because this unlinked GitHub App installation has used all three included PR scans. You have 0 of 3 included PR scans remaining. Create a free account to continue protection, manage scan settings, review security history, and control which repositories are protected. |
Member
Author
|
@greptileai please review the PR |
Member
Author
|
hey @greptileai, please re-review the PR |
Member
Author
|
hey @greptileai, please re-review the PR |
Member
Author
|
hey @greptileai, please re-review the PR |
Member
Author
|
hey @greptileai, please re-review the PR |
|
Review the following changes in direct dependencies. Learn more about Socket for GitHub.
|
Member
Author
|
hey @greptileai, please re-review the PR |
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Hold — rebuilt native timing comparisons and current-head CI are pending. Do not merge or publish.
Recording Stop inherited expensive recovery work even after clean capture: an earlier two-minute Studio case took 1.88 seconds to finalize versus 310 ms in 0.5.9, and Windows Instant took 5.36 seconds versus 427 ms after capture ended. This change uses faster in-memory byte snapshots, defers disposable staging flushes, and fuses bounded validation with remux for verified clean video. It retains packet corruption/read-error checks, seek validation, full audio/recovery validation, source snapshots, backups and final output syncing. A path-bound, single-use Instant completion token is issued only after successful encoder finalization. Temporary cleanup failure cannot turn a successfully saved recording into a failed Stop.
The PR also fixes release compilation guards and Windows DASH manifest ownership, regenerates the bindings needed to show Tauri's main window, restores GPUI editor activation and recording error display, preserves the editor destination on restart, and keeps valid 0.5.9 projects with omitted failed optional tracks usable. Mac Studio retains system-audio-before-screen shutdown acknowledgement for their shared capturer while microphone/camera stop concurrently. Other platforms keep parallel stopping.
Validation for
526dc80bf5429f7729b66d0291802b65cfeb48c7:c871d2fffhead improved the large delays but still failed latency parity: Linux 120 s Studio stopped in 695.489 ms versus 338.422 ms; Windows 12 s Instant remained 149.449 ms and 226.789 ms slower in opposite run orders. Functional decode/export checks passed. Those artifacts do not contain the latest fused-validation or stop-ordering fixes.The previous head had green main CI, A/V and CodeQL and a 5/5 Greptile PR review. A separate full-PR CLI review returned 4/5 with a generated-binding process allegation disproved by the saved canonical Rust exporter execution and matching generated blob. Those earlier review/CI results are historical. The new full-PR CLI review also verified the generated bindings and reported no remaining concern. The CLI's 727-file baseline-diff limit is honored; the complete 155-file PR is the review scope, with separate 0.5.9 source and native comparisons.
No signing approval, merge or release publication has occurred. Quiet-host Mac timing, physical-device disconnects, extended duration, Wayland/hardware variants and live upload completion remain unproven. Windows baseline playback and same-media A/V export measurements have documented failures and are not counted as passing results. Existing Mac native-stop error suppression and Tauri's upload-completion race predate the PR and are not claimed fixed. This evidence cannot guarantee zero regressions.
Greptile Summary
This PR hardens cross-platform recording completion, recovery, timing, and rendering while restoring several desktop and GPUI integration behaviors.
Confidence Score: 5/5
The PR appears safe to merge from the eligible follow-up-review findings, with no blocking failure remaining.
No blocking failure remains.
Important Files Changed
Reviews (6): Last reviewed commit: "fix: preserve saved recordings when temp..." | Re-trigger Greptile