I just threw away a bunch of tests

May 13, 2022
I had failed to complete the "refactor" part of "red/green/refactor" several weeks ago, leaving me with some redundant tests that came back to bite me.

Yesterday I just deleted a whole bunch of tests in a project I’m working on for a client.

Why? Aren’t tests supposed to be helpful?

Supposed to be, yes.

And these tests were helpful, once.

The thing is, I had written these tests while using Test-Driven Development, so I know that they had been valuable at the time they were written. So what changed?

What changed was a refactoring several weeks ago. The functionality these tests covered had been extracted to a new Go package (class in OOP parlance). When I extracted that functionality, I did it using TDD, so I wrote new tests in the new location, and the existince of the old tests in the old location served as proof that I didn’t break anything in the process.

However, I had failed to complete the “refactor” stage of “red/green/refactor”, and left the old tess in place. During my refactor, I had written a test, saw that it failed, then wrote/copied the code to make it pass. What I missed was the part of refactoring that removes duplicate code—and tests are code!

I noticed yesterday when I made a change, and it broke more than one test. Oops. The old tests and the new tests both started failing.

So I fixed the new test, and deleted the old, now-duplicate test.

One charactaristic of a good test is that a single failure will trigger a single test. It’s not always achievable, but when you see more than one test failing for the same bug, it’s definitely a smell, and something to consider. It often requires more thought than just deleting a duplicate test, but that’s an easy place to start looking!

Share this

Related Content

Which skill is more important: Testing, or debugging?

One of these skills, if you're good at it, diminishes the need for the other other.

Have unit tests ever saved you?

Three times unit tests saved me, in as many days.

Reader response: The downward spiral of manual acceptance testing

Lack of unit testing drives the need for manual testing. Since testing is bunched up, development is as well.