In addition, I see two devices with same MAC and name since yesterday. Tried to reset the device. It worked, because default names and beacons were reset too, but behaviour is still not as the first day. I think it happened after I reset it the first time.
Will try to figure out on my own though, as this is not the purpose of this forum topic.
Yes please, but something is still not right, as your selected message with the new MAC address under Theengs Gateway should now be decoded correctly.
With the PUBLISH_ADVDATA option on though we can see what might be the problem, or if the uninstall and re-install didnât work properly to pull in the new decoder version for whatever reason.
MAC list went away. Now, this is what I see in MQTT Explorer:
The undecoded topic changes randomly through many MACs.
PUBLISH_ALL: disabled PUBLISH_ADVDATA: enabled
Unfortunately, itâs beer time with my neighbour. I wonât be able to give you further feedback until later tonight or tomorrow morning, but do feel free to list actions I can do for further testings.
That is actually the correct expected behaviour, my bad for asking for it
I think it is actually just an issue with Pypi not listing the latest Decoder release yet fully in their system, as searching the Theengs Decoder still comes up with version 1.7.4
so I assume with your last uninstall and reinstall you still got version 1.7.4
Best to enjoy your beers now, and when you get back or get a chance go to pypi.org and search for theengs
If and when the search result also comes up with TheengsDecoder 1.7.5 then uninstall and reinstall the Theengs Gateway Add-on - we might have rushing it all a bit just now
Iâm not sure if you have tried again uninstalling and reinstalling the Theengs Gateway Add-on - possibly also with a HA reboot in between, if possible at all with your set-up - which might actually have been a cache issue before instead of the pypi.org listing issue!?!
The OpenMQTTGateway development branch test builds have also finished now and can be installed through
with either the esp32-m5stack-ble or a plain esp32dev-ble binary.
However, I cannot find any new Beacon-XXXXXX device under MQTT integration. Am I forgetting anything? How should I set PUBLISH_ALL and PUBLISH_ADVDATA. Tried some combinations and nothing changes.
This is the iBeacon protocol, which I read the beacon send out alternately with its own protocol, which you confirmed in your nRFConnetc screenshot and the previously undecoded message.
So expanding and looking in the history (on the right side of MQTT Explorer) of the 001D⊠topic you should see the message history, with the above stated iBeacon protocol, which we ignore, but also with the native, now decoded, messages alternating.
I donât know if you can set the beacon to only send its own protocol broadcasts, and to turn off the iBeacon broadcasts completely, as they are not relevant to how you want to use the beacon with HA.
You might want to set BT: Interval between scans and BT: Interval between active scans in the gateway interface lower from their default 55 seconds to something like 20, so that more messages will be received for this verification.
Letâs try it with just the generic esp32dev-ble binary, although from the messages it looks as if the beacon might only be sending iBeacon protocol and other protocol messages.
Before installing the esp32dev-ble binary though, can you turn on the BT: Publish Advanced Data again and see what these other messages apart from the iBeacon protocol are in the history?
If you click on the copy icon behind each message in the history you can also paste them easily in here, instead of having to upload wide screenshots
message, which we went for originally, should be correctly decoded now.
I know it might be a silly question, but you did upload the development binary from the upload page with Theengs OpenMQTTGateway DEVELOPMENT SHA:bb491a TEST ONLY
at the top, and not from the 1.7.0 release page, which looks virtually identical, but has only got the heading OpenMQTTGateway v1.7.0 ?
Great, I thought I was going mad with it not working as it should.
It does happen with the pages looking so alike
The actual device tracker AWAY time can be adjusted with the presenceawaytimer value, which is set to 120 seconds as a default after the last Bluetooth reception of the tracker has been registered, but can be adjusted. And as device-tracker entry it can also be assigned to a person in HA.
So any of the three previously mentioned ways will be possible now, Add-on only, once the updated Theengs Decoder is all fine with PyPi.org, OpenMQTTGateway only, as you are doing now, or OpenMQTTGatway undecoded binary as a proxy antenna for the Theengs Gateway Add-on.
I hope this implementation of the Feasycom beacon helps you use it the way you planned to, and thanks again for your participation in this decoder, as it will also recognise all other Bluetooth beacon models from Feasycom for other users!