Open keatontaylor opened 7 years ago
This is vastly better for absolute values than the current percentage implementation, but a potential problem is that changing the Alexa imperial/metric unit preference can cause unwanted conversion of input and output values.
For those of us on imperial units, alexa provides celsius measurements for all requests and we specify the unit in the response. So I don't see this as an issue.
@keatontaylor the API works exclusively in Celsius, we don't get to choose the unit in the response. Edit: Oh, looks like they updated the API to report units ("scale") since I did the implementation; don't think we can tell if there will be conversion performed, though.
We could provide a config variable for users to tell us which unit their alexa is set to.
That seems reasonable.
Hmm, I was hoping this could be expanded for numeric sensors too, but it seems that querying state requires saying: "Alexa, what is the 'device name' temperature"
but this still works well for input_sliders because the interactions will be limited to:
"Alexa, what is the 'input slider name' set to" and "Alexa, set the 'input slider name' to xxx"
I'm surprised that amazon hasn't released a framework for querying sensors, and setting values for devices as the temperature control framework is already so compatible. So it seems that a lot of that code in their backend could likely be reused.
My guess for the sparse API is that they are only releasing features that partners already have skills written for right now.
Been looking at the smart home skill api for temperature control and query and it seems that it is viable for setting and controlling query and setpoint control.
We can set the mode to custom and provide a friendly name for the custom mode for whatever it is we want to query.
Use cases could be querying any number of numeric sensors, changing values and querying input booleans, but I do not have any specific use cases.
@dale3h what are your thoughts on this?