API results contain a whole lot of needless boilerplate

[the-daniel-rothig]

The API result objects contain $id and $ref entries, as well as a lot of nullable fields.

The expected behaviour is that they only contain non-null entries except where it makes sense, and no random metadata injected by ASP.NET

[ashleybaldock]

The $id and $ref entries are due to serialising cyclical data structures and wanting to be able to decode them into the same structure at the other end - there's a fairly simple toggle which you can set which changes the behaviour to not do this, but then you don't get any of the cyclical references (and the data returned balloons somewhat in size).

Is this really the expected behaviour? We haven't really specced out what we want the API to do aside from be consumed by our application. Obviously all things are possible though...

Skipping null fields is plausibly easy enough to do.

[ashleybaldock]

https://abhishekcgh.wordpress.com/2014/04/26/serializing-json-data-with-cyclic-references/ - the $id/$ref are just Newtonsoft's JSON lib's way of handling this stuff (https://www.newtonsoft.com/json/help/html/PreserveObjectReferences.htm) - you can write your own deserialiser to handle this data format. Isn't ASP.NET metadata :)