[Edu-sig] Python as a first language for computer sciencist

[Kirby Urner]

> But that's probably a few years in.
> 
> Art
> 
> 

My inclination is to introduce testing pretty early, treating tests like
comments -- a good habit, supporting what in some situations is essential
practice.

Unit testing frameworks tend to be overkill though, on a first pass.  I was
thinking more in terms of doctest:

http://docs.python.org/lib/module-doctest.html

This lets us promote commenting and testing at the same time, i.e. document
your function with a couple use cases (as if this were a shell session) and
be done with it.

In general, I believe in using pre-written code, not starting from a blank
page.  This is how we learn to read human languages to:  we get finished
examples to parse, don't have to write in a vacuum.

So, a first exposure to tests will be in the form of already-working code.
Sometimes we make some changes to *intentionally break* working code.

Kirby