ciphera-net/pulse
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 15 Wiki Activity

fix: recover gracefully from stale Server Action hashes after deployment

Browse Source 
Wrap all Server Action calls (getSessionAction, exchangeAuthCode,
logoutAction) in try-catch so a cached browser bundle with old action
IDs triggers a hard reload instead of an infinite loading spinner.
This commit is contained in:
Usman Baig
2026-03-07 19:37:41 +01:00
parent cc268c320e
commit d2dfe62993
3 changed files with 25 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
CHANGELOG.md
View File

@@ -12,6 +12,7 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
### Fixed
- **Login no longer gets stuck after updates.** If you happened to have Pulse open when a new version was deployed, logging back in could get stuck on a loading screen. The app now automatically refreshes itself to pick up the latest version.
- **City and region data is now accurate.** Location data was incorrectly showing the CDN server's location (e.g. Paris, Villeurbanne) instead of the visitor's actual city. Fixed by reading the correct visitor IP header from Bunny CDN.
- **"Reset Data" now clears everything.** Previously, resetting a site's data in Settings only removed pageviews and daily stats. Uptime check history, uptime daily stats, and cached dashboard data were left behind. All collected data is now properly cleared when you reset, while your site configuration, goals, funnels, and uptime monitors are kept.