Need help with additional testing, as I only have a single device with multiple refresh rates display (240/60).

Also, changes to DisplayService.qml might be overengineered.

Here's the first issue:

• When you restart dms being on battery, it thinks that configured refresh rate is 60hz and it never goes back to the preferred 240hz

Probably can be resolved with this logic:

• Get active monitor profile refresh rate and use it as baseline
• If monitor profiles are not configured, fall back to monitor preferred refresh rate
• Re-check these steps on profile or power source change

Implemented stateless transitions between 60hz and higher refresh rates.
Also improved logging to be more human-readable.

update after latest changes: functionality looks better, but I’m still hitting one nasty crash.

I can reproduce with this flow:
plug AC -> connect external display over type-c dp -> quickshell crashes

dumps end in same path:
Qt::endPropertyUpdateGroup -> WindowsetManager::onServerCommit -> WorkspaceManager::ext_workspace_manager_v1_done (SIGSEGV)

so this looks more like quickshell workspace/output commit race on hotplug/reconfigure.
refresh-rate logic still seems to be trigger surface (mode apply happens right before crash), but probably not root cause by itself.

Resolved.

Added minimal 300ms debounce around DisplayService refresh-rate sync, and BatteryService now routes startup/power/profile/output refresh triggers through requestSync(...). This coalesces AC + external-display event bursts and avoids repeated mode applies that were triggering quickshell crash.