I was working in Storyspace 3 when, out of the blue, my Magic Trackpad 2 disconnected. With months of bitter experience of El Capitan’s Bluetooth and kernel panic problems, I thought disaster was imminent. Then the Trackpad reconnected, the pointer jolted, and disconnected again. In the space of a minute or so, I had a series of disconnections and reconnections. I plugged its USB charging cable in and restored normal control over the juddering pointer.

The Trackpad’s battery was not low by any means (77%), and to the best of my knowledge no one had opened up a heavyweight Bluetooth radio jammer nearby. Besides, my wireless keyboard had not so much as twitched through all this. In the dark days of El Cap, once the Trackpad had gone, the keyboard went too.

I left the Trackpad to charge up fully, then restarted my iMac. Once everything had settled, and without any apps open, I tried disconnecting the USB charging cable: it had exactly the same effect, a barrage of disconnections and reconnections. The keyboard didn’t flinch.

I thought it would be useful to get a log capture of what was going on when the Trackpad was trying to run wirelessly, so prepared a capture file in Terminal with
touch ~/Documents/logbtprobs.text
I then disconnected the USB charging cable again, and tried to use the Trackpad through the resulting barrage. After a minute or so, I had had enough (and I suspect the Trackpad had too), so I reconnected its cable, and entered the following in Terminal
log show --style syslog --info --last 2m | cut -c 1-22,43-999 > ~/Documents/logbtprobs.text
(That works around the bugs in log show and gave me an excellent and complete record of log entries for those two minutes.)

The first minute of that period was fairly quiet and uneventful. Then came the first disconnect:
2016-10-04 16:53:42.31 kernel[0]: (AppleMultitouchDriver) [HID] [MT] AppleMultitouchDevice::willTerminate entered
2016-10-04 16:53:42.31 kernel[0]: (AppleMultitouchDriver) AppleActuatorHIDEventDriver: message service is terminated
2016-10-04 16:53:42.31 kernel[0]: (AppleMultitouchDriver) AppleActuatorDeviceUserClient::stop Entered
2016-10-04 16:53:42.31 kernel[0]: (AppleMultitouchDriver) AppleActuatorDevice::stop Entered
2016-10-04 16:53:42.31 kernel[0]: (AppleMultitouchDriver) [HID] [MT] AppleMultitouchDevice::stop entered
2016-10-04 16:53:42.31 kernel[0]: (AppleMultitouchDriver) AppleActuatorHIDEventDriver: stop
and a reconnect
2016-10-04 16:53:42.47 AirPlayXPCHelper[138]: (BluetoothAudio) [com.apple.bluetooth.BTFigE] Howard Oakley’s Trackpad connected
and a disconnect
2016-10-04 16:53:48.79 kernel[0]: (AppleMultitouchDriver) [HID] [MT] AppleMultitouchDevice::willTerminate entered
2016-10-04 16:53:48.79 kernel[0]: (AppleMultitouchDriver) AppleActuatorHIDEventDriver: message service is terminated
2016-10-04 16:53:48.79 kernel[0]: (AppleMultitouchDriver) AppleActuatorDeviceUserClient::stop Entered
2016-10-04 16:53:48.79 kernel[0]: (AppleMultitouchDriver) AppleActuatorDevice::stop Entered
2016-10-04 16:53:48.79 kernel[0]: (AppleMultitouchDriver) [HID] [MT] AppleMultitouchDevice::stop entered
2016-10-04 16:53:48.79 kernel[0]: (AppleMultitouchDriver) AppleActuatorHIDEventDriver: stop
and a reconnect
2016-10-04 16:53:50.33 AirPlayXPCHelper[138]: (BluetoothAudio) [com.apple.bluetooth.BTFigE] Howard Oakley’s Trackpad connected
and so on.

There was no sign of anything which precipitated these – no crashing components or apps – so I think it is fair to conclude that these were initiated by the Trackpad itself. In other words, the Trackpad is broken. I haven’t yet tried the traditional remedy of last resort, turn it off and back on again, nor have a tried using the Trackpad with my MacBook Air, to see if it has the same problems with that. I will try those a bit later.

It works fine tethered via its USB charging cable, which is a comforting fallback, but clearly not my intended solution. I suspect that I will be taking this Trackpad on a ferry journey for a visit to the Genius Bar at the Apple store in West Quay, Southampton, in the near future.

It would – just – have been possible to do this log examination using Console 1.0, but given its very limited functionality, it would be far from ideal for the task. The Bluetooth tools supplied with Xcode would have been of little use to such an intermittently-connected device. I just wonder how on earth I can talk a MacFormat reader through this sort of process by email, when I try to tackle their questions.

The good news is that, so far, Sierra’s Bluetooth drivers and kernel seem to cope with such insults without collapsing in a snotty heap, which is more than could be said for El Capitan’s.