v0.9.1
0.9.1 — 2026-08-01
Mostly monitoring. It used to be a small dialog; it’s now a proper part of the app.
Monitoring
- A full surface instead of a 640px dialog. Navigation lives in the sidebar, and the panes keep their state — leave and come back and you’re where you left off.
- The dashboard is a ranked worklist. What needs attention sits at the top, and rows expand in place. The separate Live pane is gone; the dashboard does its job.
- Sources are configured in one place, showing cadence, last-checked, failure counts, and stale enrolments.
- Four explicit alert types, including active backlog and oldest-message age — not just dead-letter counts.
- Fire only on a sustained breach. A for-duration qualifier stops a single spike from waking you.
- Create an alert from where you already are — a queue’s tab, the sidebar menu, or a dashboard row. One click starts monitoring the active subscription.
- Clicking a notification opens viewbus on the alerting resource (Windows), and alerts can re-notify while they stay firing.
- History and trends. 24-hour sparklines, dead-letter trends per estate and namespace, MTTR and flappiest/longest-burning rankings, and CSV export.
Navigation
- Back and forward. Mouse side buttons and Alt+←/→ retrace your steps; Esc leaves monitoring.
Fixes
- The message list no longer shows messages that are gone. If a message dead-letters while you have it open, the counts refresh — and now the list and the detail pane follow, instead of leaving you reading something that has already moved.
- Manual connections survive an unreadable keychain instead of vanishing from the sidebar.
- Concurrent database writes wait their turn rather than failing outright.
- Webhooks refuse to send unsigned rather than falling back silently.
All local-first as always. Update via the in-app updater or grab a fresh installer from viewbus.app — your settings, themes, monitored sets, and stored connections carry over untouched.