This week I had a conversation about writing tests with my boss and another coworker. When I first onboarded, our company had seven Rails developers working on 2-3 projects simultaneously. We had some request specs to test an API for one project.

A couple months later I was tasked with integrating a payment service. I started with writing unit tests to help me guide the design of the service. It wasn’t a complicated object to test, just a couple public methods to massage the data payload and a callback function to process the bank response. While writing tests by definition tests the functionality of the system under test, I prefer to emphasize the overall positive impact it had on how I went about designing the service. Having the benefit of being confident about new code not breaking existing code was a nice add-on.

Fast forward to the present where our team is now six Rails developers working/maintaining 10+ projects there has been more bugs and burden. I can’t say I understand how our company manages projects and schedules, but I do know that it is vastly different from what I’ve experienced back in my days of being a fed.

Cost, schedule, performance. Pick any two. That has always been the mantra for project management. With overtime pay rarely being enforced in Taiwan, the two prioritized factors are always cost and schedule. I ain’t gonna front, I haven’t written tests for several entire projects. The rest of our Rails team hasn’t either.

I feel the project management philosophy is strange. At times I am bamboozled, bufuddled, and confused. Are we really just gonna gamble and hope the end user doesn’t use this particular functionality or interact with the system in a certain way?

I’m certain there is some shady stuff going on behind the scenes. What is our mission? How can we have more projects than engineers? It’s 2018 and cash still rules everything around me.

In any case, straight up jumping into TDD with little or no prior experience is something one simply does not do. There is a learning curve. Our current waterfall approach to system development is not cutting it. We haven’t written enough mature libraries to be at that point yet. The story goes on and on.