Agreed. I'd argue that if a large number of tests break when a "large" change occurs in an application, those tests are either testing the wrong thing, or writt

Agreed. I'd argue that if a large number of tests break when a "large" change occurs in an application, those tests are either testing the wrong thing, or written with bad starting assumptions. Yes, there's a transitional period where the responsibility may shift from the caller to the call-ee to take care of some resource, and yes, stuff will break, but the amount of test breakage should be proportional to the impact to the application. I greatly-prefer the warm fuzziness of knowing that I have a large number of unit tests as a "safety net" to detect if my own thoughts on what is "correct" are in some way wrong.