Ideology – do unit tests make type systems unnecessary? (2015)

> simple unit tests aren't the important ones

And yet, they prevented me from committing dumb mistakes to the main branch so many times.

> types are just simple unit tests written for you

Even better if I don't have to write them. Is this really a counter argument?

> dynamic languages only need unit tests because they don't have type systems

But every language has a type system! How would they differentiate an integer from a string if they don't?

Even assembly has a type system (byte vs word vs double-word).

Am I just nitpicking or is that argument fundamentally flawed and ignorant?

> What's going on here? These can't both be right.

They can both be wrong though, and they seem to be IMHO.

This does not incite me to watch the 20min video. This seems too click-baity.

This sounds a bit like a 'no true Scotsman' (i.e. the bar would be moved with any example of a bad unit test that needs types), but I would need think more on it.

In my opinion, unit tests need to cover multiple levels, at least two, the units, and also the business logic, and proper types would cover the gaps in between. If anything I would invert the title of this, and say a strong enough type system makes unit tests unnecessary, but I don't think that's anything new and you can see it at work in pure FP languages like Haskell.

Some people claim that unit tests make type systems unnecessary: "types are just simple unit tests written for you, and simple unit tests aren't the important ones". Other people claim that type systems make unit tests unnecessary: "dynamic languages only need unit tests because they don't have type systems." What's going on here? These can't both be right. We'll use this example and a couple others to explore the unknown beliefs that structure our understanding of the world.

I think "types" are just compile-time unit-tests.

But unlike conventional "unit-tests" they also apply inside the code, not just on its externally observable call-boundary.

So perhaps more appropriately then types are "reusable assertions".