Distinguish disabled routes from monitoring-off routes in status text
This commit is contained in:
@@ -270,3 +270,5 @@ Roughly in priority order:
|
||||
`v0.16.72` fixes the favicon not appearing in some Chromium-family browsers (reported: DuckDuckGo's browser showed the default globe icon while Safari showed the real one correctly). The app previously declared a single `<link rel="icon">` pointing straight at the 1018x1001, 396 KB source PNG with no `sizes` attribute and no `/favicon.ico` fallback -- Safari is forgiving about oversized, unsized favicons, but some Chromium-based browsers silently skip one that large rather than downscale it, and several also probe `/favicon.ico` directly regardless of what the `<link>` tag says. Added properly sized `favicon.ico` (16/32px, multi-size), standalone `favicon-16.png`/`favicon-32.png` with `sizes` attributes, and a 180x180 `apple-touch-icon.png`, all generated from the existing approved icon artwork and served automatically by the existing static file handler -- no server route or icon design changes.
|
||||
|
||||
`v0.16.73` makes the Dashboard's System panel top accent bar reflect resource state, instead of always showing green -- pointed out after a screenshot showed CPU pinned at 100% (its own stat correctly shown in red) while the panel's top bar stayed the hardcoded `var(--green)` it always had. `renderHeroPanel()` now tracks the worst tone across CPU/memory/swap/disk (the same "warning" at 75%+ / "critical" at 90%+ thresholds each stat's own value and fill bar already used) and applies a `tone-warning`/`tone-critical` class to the panel, which `.system-panel::before` now reads instead of a fixed color. Network, uptime, and throughput don't carry a tone and are excluded from the calculation, same as before. Applies to the Dashboard's System panel only -- the Administration > System tab's equivalent hero grid has no top accent bar to react.
|
||||
|
||||
`v0.16.74` fixes an inaccurate status label on Hosted Site, Proxy Host, and Streaming Host cards, and the Certificates table's Upstream column: a route with monitoring intentionally turned off via "Monitor this site/upstream" in Advanced options -- while the route itself stays enabled and running -- read "Monitoring paused", the same text used for a route that's fully disabled. "Paused" implies a temporary interruption; deliberately unchecking the monitor box is an ongoing, intentional setting. The backend already distinguished the two cases (`status: "disabled"` when the route itself is off vs. `status: "unmonitored"` when only health checks are off, in `checkProxy()`), the frontend just collapsed them into one string in four places. Disabled routes keep "Monitoring paused"; a running route with health checks off now reads "Monitoring disabled". No backend or status-logic changes, text only.
|
||||
|
||||
Reference in New Issue
Block a user