Passing an invalid key to a data relation column doesn’t throw an error? It just puts a NULL in the column?
Is this somewhere a setting?
Workaround is a business logic layer but I was not expecting this behaviour, that’s why you have a data relation column (foreign keys in a relational DB).
Could you please demonstrate with a sample rest request what you are doing?
Hi Mark,
I looked at my tests again and I used the wrong objectid, I fixed this but now I get a strange error:
java.util.HashMap<java.lang.Object,java.lang.Object> out of START_ARRAY token\n at [Source: UNKNOWN; line: -1, column: -1]”To add a passenger to a car, you need to use the API that establishes a relation. The REST Console doesn’t provide UI for that functionality. The documentation for the API call is available here:
https://backendless.com/docs/rest/data_set_add_relation_rest.html
“Both parent and child objects must exist in the Backendless storage.” + call to update the relation.
So 2 call for one record, better in one business logic call?
About business logic calls: if the business logic is just one SQL (*Advanced Object Retrieval"), how is the performance compared to a Database API with the same SQL?
If I understand your question correctly:
Through the business logic, you essentially use the same API requests, + the business logic itself. Therefore, normal use of the API will be more productive.
It will be more convenient for you to use the API for Deep Save:
https://backendless.com/docs/rest/data_deepsaveoverview.html
Advanced Object Retrieval:
https://backendless.com/docs/rest/data_general_api.html|
Best Regards, Nikita.
I was wondering how much delay the blockly way of working is, are the business logics compiled or everytime interpreted?
Business logic uses JS. It is interpretable.
Thx for the info! 
To clarify, business logic written with BLOCKS uses JS.
You can also write it in another language. Depending on this, it may be compiled.
Best Regards, Nikita.
