Promises Coverage

Fleeting

On the contrary to code coverage that tells how much of the code is covered by the tests, I thrive for considering how much of the promises are covered.

code coverage often falls into the measure-objective inversion problem and also gives a false sense of truth.

100% coverage does not mean anything from the user experience perspective. Worse, it hides the fact that maybe a lot of those 100% are just dead code never used, but only tested.

Notes linking here

• place of tests in software development
• unintended consequences of code coverage
• verified/proved promised contracts based programming

Permalink