Repository navigation
fix(desktop): preserve causal timing for typing before click presentation - #112
Merged
Merged
Conversation
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Fast controlled typing can begin before the focus click's first frame is submitted. The visual probe previously discarded these early keys, then matched their counter responses to later presses, underestimating latency before failing correlation. Retain provisional keys behind the anchored pending click in the same ordered counter sequence; presentation still confirms the click before arming keyboard measurement.
Three regressions cover delayed click presentation, a collapsed 32-key response with focus cancellation, and an unanswered click timeout. The existing pending bounds, modifier and surface checks remain. This only corrects opt-in diagnostics and does not claim to fix physical network stalls or change actual input dispatch.
Validation: the initial Mac desktop suite passed112unit tests and all integrations; the final11visual-probe regressions include the added timeout case. Final fmt and strict workspace desktop Clippy pass. The first workspace run reached doctests but failed while a different feature build used the same target; the separate uncontended workspace doctest rerun passes (shared-target interference is an inference). Optimized Mac build and both native artifact CI lanes pass; CodeQL passes. All16required check results passed (with two normal skips); PR112 merged as a3fd24b. Corrected installed-viewer visual timing remains pending. Details: docs/reports/rds-visual-probe-focus-race-20261006.md.