It’s a stopgap right now to make sure that retained messages are cleared. Some of the brokers I’ve tested do not overwrite retained messages correctly.
I had not seen that. Interesting. From my own testing with Bluetooth over the years, though, I can say that the OP is getting very far ahead of him/herself by suggesting meter accuracy and reliable room-level triangulation. Far to much environmental variability.
There’s a good reason that Apple’s iBeacon spec and Google’s Eddystone spec are granular with distance predictions only to “near” “far” “very far” and the like. Distance measurements from RSSI are not consistent or reliable.
Having this access to GH’s hardware is very interesting though.
Hi, Can anyone post their “behaviour_preferences” for ‘presence’ as at the moment this is killing my network?
Cheers to everyone contributing to this & that of course includes you @andrewjfreyer
Looking forward to trying out the motion’ script & was wondering in its present state would it be worth while moving over now from presence or waiting?
After a device was identified, a line with topic / payload is printed in the log. On each subsequent lines I have
Error: Invalid port given: 0
Use ‘mosquitto_pub --help’ to see usage.
Only arrive / depart mqtt messages are received by my broker. To me it looks like there is something preventing mosquitto_pub getting called correctly. In publish_cooperative_scan_message() the port argument is not given, but this does not seem to be the cause.
Incorporated the feature, but without a merge. Wanted to maintain consistency through other parts of the (stupidly massive, now) script. Please let me know if this is suitable for your purpose. Note that the variable name in the mqtt preferences file is different from the pull request.
Not sure if this has been asked over the 442 posts, sorry if it was.
Would it be possible to use this with a OBD car adapter to detect if a car is home?