Background: Why Scalatron?

[Scalatron]

Just in case you're interested, here is some background information on how

Scalatron came about, extracted from a private e-mail related to the topic. It

explains some of design choices in the first version:

At the moment (version 0.9.2, March 2012), Scalatron is an early and at the

moment relatively simple project. I created it for a little workshop I was going

to organize for Java-programming colleagues and friends that I wanted to get

into Scala. The first impulse was obviously to do what everyone does: find or

create a PowerPoint presentation and walk people through features, maybe with

some live coding, either (like I've seen it done at some of the user group

meetings) with the presenter coding on screen and them following along or with

them typing in examples based on something projected up front. But my prior

experience with this is that while people may listen attentively, they are

generally not fully engaged the whole time and - most importantly - unless they

discover a solution and perform an action on their own, it won't stick. They'll

go home and most of what they heard will evaporate and not be actionable for

them.

So I thought I should maybe rethink the whole "training" approach and ask about

the goal first: what is the ultimate goal of such a workshop? Obviously it is to

get people productive in the new language within some finite time, in my case a

single six or seven hour session. The primary levers here seemed to be to (a)

get people to do a lot of hands-on coding, (b) maximize the motivation to make

progress, (c) provide immediate and frequent gratification and (d) give everyone

the ability to move along the steepest learning curve their prior experience and

analytical gift permits.

Letting people write simple bots for a simple game, based on a tutorial that

walked them through an increasingly complex series of bots while introducing and

explaining the basic features of the language in a carefully chosen sequence

seemed like a better answer. I was also hoping for people's competitive

instincts to propel them through the material at the fastest possible pace;

after all, if your bot is competing with everyone else's on a projector that

everyone can see, you want to do well.

The Scalatron package was designed around these basic ideas, although there were

interesting alternatives at some points, which maybe could be explored in a

future version. But for the first version, simplicity was important, both

because my own resources are finite and because it limits what needs to be

explained. So that's where Scalatron comes from.