Conversation
This reverts commit c12ff7f.
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 95a23d55ea
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| } | ||
| } else if (parsed.session?.user?.userId) { | ||
| this.sessions.set(parsed.session.user.userId, { | ||
| if (parsed.session) { |
There was a problem hiding this comment.
Handle legacy auth-state format during restore
initialize() now restores only parsed.session, so auth files written by the reverted multi-account build (sessions + activeUserId) are ignored and this.session stays null, which silently logs users out after upgrading to this commit. Please keep a migration path that reads the legacy sessions payload and promotes the active entry into session before persisting.
Useful? React with 👍 / 👎.
| this.escapeTapDispatchedForCurrentHold = false; | ||
| this.clearEscapeAutoKeyUpTimer(); | ||
| // Start the hold timer (will be cleared on keyup if released before 5s) | ||
| this.startEscapeHoldRelease(pointerLockTarget); |
There was a problem hiding this comment.
Schedule Escape keyup fallback when starting hold logic
In the pointer-lock Escape path, keydown starts the 5s hold timer and returns early, but no fallback keyup is scheduled; if the browser suppresses Escape keyup while keyboard lock is active, the timer is never canceled and a short Escape tap is misinterpreted as a hold, causing unintended pointer-lock release after 5s. scheduleEscapeAutoKeyUp() exists for this case but is never invoked here.
Useful? React with 👍 / 👎.
Reverts #356