Skip to content

PebbleKit2 messages from watch #11

@petrnalevka

Description

@petrnalevka

Hello,

I have some issues with PebbleKit2 communication watch->phone.. From phone->watch everything works well.

But sometimes non-deterministically my messages from Pebble to Phone are not delivered into the ListenerService.

I have debugged the watch app and I see the message is sent from the watch and acked by the Pebble app:
10:24:15] SleepAsPebble.c:372> Action message: 4
[10:24:15] SleepAsPebble.c:655> Data acked successfully.

I see the messge rrived at Pebble app on the Phone:
2026-03-06 10:38:04.780 17501-10189 PebbleProt...:D5:83:90) coredevices.coreapp D sending io.rebble.libpebblecommon.packets.AppMessage$AppMessageACK@fffffff2
2026-03-06 10:38:07.919 17501-10189 PebbleProt...:D5:83:90) coredevices.coreapp D inbound pebble protocol packet: io.rebble.libpebblecommon.packets.AppMessage$AppMessageACK@ffffffef
2026-03-06 10:38:18.211 17501-7431 PebbleProt...:D5:83:90) coredevices.coreapp D inbound pebble protocol packet: io.rebble.libpebblecommon.packets.AppMessage$AppMessageACK@fffffff0

But the override suspend fun onMessageReceived(
watchappUUID: UUID,
data: PebbleDictionary,
watch: WatchIdentifier
): ReceiveResult {
method in my app is not called.

My Pebble app has the comapnion app section in package.json
"companionApp": {
"android": {
"url": "https://play.google.com/store/apps/details?id=com.urbandroid.sleep",
"apps": [
{
"package": "com.urbandroid.sleep"
}
]
}
},

And this is the manifest declaration of the service:




Interestimgly the onStartApp method is called but not onAppStop or onMessageReceived

After some debugging I have also found this error in the logcat:
026-03-06 10:48:35.711 17501-13801 e coredevices.coreapp E Unhandled exception in PebbleKit2 24b0a1e5-f0b1-440d-8e53-8f26688a3f07: All 2 AppMessage channels already claimed
java.lang.IllegalStateException: All 2 AppMessage channels already claimed
which is probbaly the cause of the problem...

I seen that force closing the Pebble app fixes the issue.

Could this be cause by not closing the sender properly? I have improved sender closing handling to always close X seconds after the last successful message. Is this a good approach?

I guess closing the sender just afetr sending the message and creating a new one for each messge is not a good approach - I guess this woudl break the ack mechanism?

The github readme could be a little more specific on how to best handle sender closing IMHO..

Many thanks for all you effort on this!

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions