Sprint Planner Helper – Session 7

Posted by Marcus Hammarberg on February 11, 2009
Stats

I’ve been doing some major reworking of the repository that I am using, from being a product repository to being a ProductOwnerRepository and a ProductBacklogRepository. This took me more than one hour and therefore I did it as a prolonged check-out session.

Halfway through that prolonged session I realized that I still don’t quite get the repository pattern (by the way, here is a great post about the repository pattern) or rather the way it’s being used in the ASP.NET MVC Storefront Starter Kit. The thing that had me confused is the service… It feel like an extra layer on top of the repository that I don’t understand.

But 06:03 into part 8 of the video series I got it… What Rob Conery is doing is to overload the constructor of the controller so that he can send in a fake version of the repository. Then the controller actions are creating the services, which take a repository interface as constructor parameter (i.e. dependency injection). For me that is not “testing only one thing”.

I am wondering if not the services would be better of implementing a interface that can be mocked from the tests of the controller. In that way you will only test out the controller and service functionality separately.

OK – but what is a service then. Right now my repository and service contain the same functionality, which feels quite strange. From the definition on wikipedia we learn:

Services: When an operation does not conceptually belong to any object. Following the natural contours of the problem, you can implement these operations in services.

So I guess I was a bit fooled by my years in layered architectures where there is a strict hierarchy. A method that get something from a database doesn’t sit well in any entity and can therefore be placed in a service that get it from a repository. Once the service returns the entity it takes responsibility for all business rules. At least that’s how I understand it now.

I am still loving the way TDD is driving my design-decisions. For a long time i pronounced TDD Test Driven Development but as you use it you realize that the last D is Design. You actually learn quite a lot about the code as you write your test and do small design changes as you go along.

On good thing that I learned in my last project is to create a class that hold test data. This time I actually test-drove (?) the content of the data. So I have test that test my test-data… A bit strange maybe – but it work out nicely.

I also took the time to rename my blog posts to number them 1..n instead of keeping track of which hour I am on… Felt to much like reporting time as consultant.



Published by Marcus Hammarberg on Last updated