Question

This is “what is the best practice” question.

I have 3 tables: orders, customer and address.

Order
------
ID
ADDRESS_ID


Customer
------
ID


Address
------
ID
CUSTOMER_ID
IS_HIDDEN

User can have multiple address and is free to edit them at any time. When user places an order I need to store a “snapshot” of his address at that time, associate it with the order and make it unavailable to the user. In other words, I want to duplicate address row which will be owned by Order table instead of Customer. This ownership needs to be somehow indicated (for example with IS_HIDDEN flag).

Is this correct approach that some data from address table is sometimes “owned” by a Customer table and sometimes by “Order”?.

The other solution would be replicating all columns from Address table in Order so the ownership will be explicit but I have bad feelings about duplication in schema.

Answer 1

If you really want a snapshot of the address you should probably store the address in the order, not a separate table.

You would be making your queries and data unnecessarily complicated.

And with the implementation you are describing you would get duplication anyway.

Example:

Customer Address = 1 street

Customer places order the address is hidden in the table.

Customer places another order with the same address. The address is
saved and hidden in the table.

The final result would be:

Customer ID | Address  | Hidden  
    1       | 1 street |  false  
    1       | 1 street |  true  
    1       | 1 street |  true