[gdevenyi]

Summary

As agreed on #580: a headset on its charging cable now counts as connected.

 checkDeviceConnected()
   r = getBattery()
-  return ok(r) && r.status == BATTERY_AVAILABLE
+  return ok(r) && (r.status == BATTERY_AVAILABLE || r.status == BATTERY_CHARGING)

The other three statuses still mean disconnected, since each one means we couldn't talk to the headset: BATTERY_UNAVAILABLE (off or not paired), BATTERY_HIDERROR, BATTERY_TIMEOUT.

Evidence

Tested on a Virtuoso XT with the cable and the receiver both plugged in. This build includes #570's battery code, because master's Virtuoso code never reports charging.

• Before: wired 0a62, BATTERY_CHARGING → --connected prints false
• After: wired 0a62, BATTERY_CHARGING → true; receiver 0a64 (no headset paired) → still false

The unit suite passes and the file is clean under clang-format 18. checkDeviceConnected() short-circuits for the test device, so no unit test reaches this line.

Merge Danger

Door: two-way. It's a one-line revert.

Blast Radius: behavioural

This changes what --connected prints for all 18 device files that can report BATTERY_CHARGING. A script that used --connected as a "not charging" check will now get true while charging. It wants a line in the release notes, and the PR title is written to serve as that line.

Checklist

• I adjusted the README (if needed). Not needed: --connected isn't documented beyond --help.
• For new features in HeadsetControl: I discussed it beforehand in Issues or Discussions. Discussed on Add CAP_LIGHT_COLOR (--light-color RRGGBB) #580.

🤖 Generated with Claude Code