/learning response change

[sjsafranek]

it would be nice to get a json response from learning like the following
{
"message": "blah",
"details": {"name":"blah", "APS": 10, "user":"me!"},
"success": true
}

this would make it easier to read with programs utilizing the find system

[schollz]

Whats "name"?
Also why do you need "user"? User should be given to the server, so the client will probably know about it.

[sjsafranek]

The Username is in the current message that is returned (it shouldn't be needed). Name is location name.

[schollz]

Hmm. So the purpose of "message" is to relay the Debug message that the server itself into its own log. This is useful to me, as a developer, cause I can use a client to "see" the server Debug messages, since currently the App prints out the data['message'].

For practical functionality though - does the client need any of this? It already knows the username and the location and the number of APs. The client doesn't even have to output a 'message', it only does right now because its useful for my debugging purposes. Maybe the client only needs to respond to success (which should probably change to an HTTP status code).

[sjsafranek]

The client probably doesn't need to see any of this. However I imagine for writing API tests it would easier to return the location name, and user within a more "machine" friendly message. If an incorrect user is returned this could highlight major async bugs that could potentially occur.