Question about syllabus design

Exercism Building Exercism
1

Let’s start from the assumption that the author of a new syllabus can write halfway decent concept docs, but is largely useless at thinking up new concept exercises (asking for a friend…)

It seems there are two ways to go:

  • Write a relatively long, multi-section concept, paired 1:1 with the concept exercise.
  • Write multiple shorter concepts, but have them all share the same exercise.

Essentially, this affects the number of boxes appearing on the syllabus tree. I’m pretty sure I’ve seen other tracks that use both approaches. For example, Go has separate range-iteration and type-definitions concepts, but they share chessboard as an exercise.

Which do we, collectively, think is better?

FYI, this is something I’m dealing with as I try to write up the (very complex, very important) Julia type system. There aren’t many existing concept exercises suitable for me to steal.

3 Likes
2

I prefer 1:1 as much as possible, and find Go’s syllabus a tad frustrating/confusing since once you finish the exercise, it’s difficult to see the “other” related concepts.

But that’s just me. I think it’s fine to do a many → 1 concepts to exercises provided there aren’t new techniques introduced with the new concepts. Otherwise, you get this mismatch between the earlier concepts and the later ones…and all of it needs to be described in the exercise introduction docs anyways.

4 Likes
3

Thanks for that, Bethany. I also have an instinctive preference for 1:1, and the 18 concepts so far merged for Julia all work that way.

I’m just looking for an escape route in case I get stuck!

1 Like