Replies: 3 comments 1 reply
-
Sorry, I think you lost me. There should not have been any changes to the behaviour of the Have I understood this correctly? I believe the behaviour has always been to send an empty string - which can sometimes be displayed as "(null)" - when the car is not being driven. We have to send something, because we set retain=true, and if we don't send something the old state will remain current. And it doesn't maker sense for a client to receive a shift state when the car is not being driven. I don't think there were any changes to this code before 1.28.5, although I refactored the code in recent master (not yet released). |
Beta Was this translation helpful? Give feedback.
-
Yes, I searched my old logs It is permissible not to send anything. It's not Teslamate but everything else that needs to change. This can be done here. It's been dealt with. |
Beta Was this translation helpful? Give feedback.
-
I can't read that from the log file. :(
Of course that was my mistake because I didn't pay attention to the payloads and just continued working with the message pointer. Then I noticed that Homeassistant (python) also reported errors more often.
I don't think it's a bug in Teslamate. Maybe something just changed in the elixir. only the Teslamate Doku writes
|
Beta Was this translation helpful? Give feedback.
-
like "teslamate/ns2/cars/2/shift_state / (null)" ......
This was different until 1.28.5 and can confuse post-processing (openhab, Homeassistant, Arduino MQTT, etc.).
Wouldn't it be better not to send such messages in the first place?
Beta Was this translation helpful? Give feedback.
All reactions