Julia Evans

Why a C++ programmer might say "I could never write Python"

I once heard a C++ programmer say “oh no, Python, that’s so hard, I could never write that. I’m so used to having a compiler!“. And at the time I thought they were just being silly – Python isn’t that hard! We’re humans, we can learn and grow! Of course they could!

But I’ve been writing both Scala and Ruby lately, and I’ve had some new thoughts. There’s not much to them, and maybe they’re obvious, but:

If you work with a compiled typed language like Scala, you develop a lot of skills around working with the type system to get better correctness.

If you spend all your time working in Python, by default you can’t even detect basic typos in your code like

def foo(elephant):
    return elephant + 2

So you need to spend all your time learning how to write correct code without a static type checker, partly by writing a much better test suite, by using a linter, etc. Tests that wouldn’t tell you very much at all in Scala (that just run the code and don’t check the result) suddenly become incredibly useful! And it’s extra important to write testable code.

So maybe the C++ programmer who says she can’t write Python is really saying “Writing safer code in a dynamic language is a skill that takes time to learn! I have not yet learned it! I would be too scared to commit to writing a reliable Python program right now!”

And maybe this is part of why Haskell programmers get so attached to Haskell – because they’ve invested so much in learning the type system, and those skills don’t transfer well to other languages like Python.

I’d be interested to know what you think. (though: I do not want to talk about whether {Python, Ruby, Javascript} are better or worse than {Scala, C++, Haskell}. There are already too many flamewars discussing that so we’re not going to talk about it. I just want to talk about skills attached to specific kinds of programming languages and how transferable they are.)

Data Day Texas 2015 A/A testing