I have observed quite a few articles lately that elaborate on issues with TDD. Nevertheless, they focused on the first letter but miss the focus of the other two letters.
Not A Testing Strategy
If you take anything out of this article, please think about this quote:
If TDD was about testing it would have been called TDT (test driven testing).
The fact that we do test upfront in TDD does not mean at all that there is a direct relationship with a testing strategy, and as many preach, unit testing is not enough to create robust software.
A Design Strategy
TDD is actually a Design Strategy, this is why the 2 last letter are for driven development. This means that your final code is being moved by those tests and not the other way around.
The design that TDD will move you towards to is minimalistic. Reducing the tendency of overengineering solutions when you don't need them. This brings a reducing time to market, by reducing the accidental complexity.
When doing TDD most developers have the complexity of letting go their egos, the problem when people fight against the practices is because they think to know better. Nevertheless, it tends to generate waste because most code optimizations tend to be premature and most extensibility points will never be modified.
There are places where TDD does not fit, for example while investigating a technology through a spike or PoC because in these cases, the person is exploring knowledge not generating value. In other cases, TDD allows you to bring value in the shortest way possible.
Conclusion
If you are an experienced developer, do not discard TDD because you think you know better, allow it to challenge you. If you are a new developer, learn from the different ways of doing things and understand the value, don't take articles at face value.
Top comments (0)