Minor fixes found on the first live production bambu test
Fixed
Unassigning a spool sent duplicate requests
Choosing "Empty" in a toolhead's spool dropdown fired one unassign request for every time that dropdown had been opened, so a dropdown opened three times sent three requests on a single click. Rebuilding the spool list was reusing the same "Empty" element while attaching a fresh click handler to it each time, so the handlers piled up. Nothing was actually harmful, the extra requests found nothing left to unmap, but it was wasted round trips and it made the logs look like the spool had been unmapped several times over.
Log noise from no-op unmaps
FilaBridge logged "Unmapped <printer> toolhead <n>" even when that toolhead was already empty. It now only logs an unmap that removed a real mapping, which is what made the duplicate requests above look worse than they were.
Both of these turned up in logs from the first live Bambu A1 test run, thanks to the developer preview.