[#8953] Outbound HTTP proxy configuration — A new Proxy tab on the Appliance page lets appliances on restricted corporate networks route their outbound traffic — license activation, agent binary downloads, and Docker image pulls, plus optionally API test and uptime check execution — through a corporate HTTP proxy, configured once with scheme, host, and port.
[#8960] Idle skip in SessionSight Replay — Idle periods are now clearly visible on the replay scrubber (including over the already-played portion), and during playback the player shows a "User was idle" overlay and automatically skips ahead to the next recorded activity.
[#8954] Select all in Sync Collections — The collection editor's selection summary now offers "Select all" alongside the existing "Clear all", so building a collection with every entity no longer means ticking each item by hand.
Improved the Docker and Updates tabs with banners that name exactly which components each tab manages (service containers vs host-installed components) and cross-reference each other, so updating one half no longer silently leaves the other stale. [#8685]
Improved the Sync → Repositories help to explain that Master mode is a deployment-time setting of the target appliance, so the "Not a Master appliance" message no longer leaves you searching for a toggle that does not exist. [#8955]
Improved the API Testing datasource dialog to offer only the supported backends — and the zero-configuration JSON backend, which previously could not be selected at all, is now available and the default.
Improved the Datasources page so freshly provisioned, not-yet-referenced datasources read "No entities yet" (with an explanatory tooltip) instead of a misleading "Not used", and module wizards now pre-select the only matching datasource automatically. [#8965]
Improved the SSO sign-in error for blocked accounts to explain what happened and that an administrator can restore access. [#8951]
Fixed the per-check Enable/Disable toggle on the Uptime dashboard and checks list silently doing nothing; toggling a check now actually changes its state. [#8957]
Fixed the SessionSight Replay progress bar hanging and Play appearing dead on large recordings, including replays not restarting after playing to the end. [#8963]
Fixed SessionSight Replay duplicating recorded page elements in the viewport. [#8959]
Fixed the SessionSight Replay pane being scrollable by the viewer; the replay now behaves like a video, driven only by the recorded user's scrolling. [#8961]
Fixed the SLO wizard crashing with an unhandled error while typing in the Uptime check IDs field. [#8964]
Fixed the managed script editor collapsing to a single line, making scripts impossible to view or edit. [#8925]
Fixed the Services Configuration Wizard failing at Apply with a Grafana communication error because Grafana was still starting; the wizard now waits for Grafana to be ready, just as it already does for InfluxDB. [#8842]
Fixed the E2E Monitoring trend chart silently freezing after the page had been open for a while, even though the rest of the dashboard stayed live. [#8952]
Fixed Service Levels datasources created from the copied internal connection string ending up UNREACHABLE; the dialog now defaults to the supported "Use local PostgreSQL" path. [#8947]
Fixed imported API Testing flows rendering without the connecting lines between nodes. [#8969]
Fixed the support ticket success panel exposing an internal file name, and removed internal work-item links from the ticket detail view. [#8481]