The Wiert Corner – irregular stream of stuff

Jeroen W. Pluimers on .NET, C#, Delphi, databases, and personal interests

  • My badges

  • Twitter Updates

  • My Flickr Stream

  • Pages

  • All categories

  • Enter your email address to subscribe to this blog and receive notifications of new posts by email.

    Join 1,862 other subscribers

Structure and Interpretation of Test Cases • Kevlin Henney • GOTO 2022 – YouTube (hello Leap Years!)

Posted by jpluimers on 2025/12/30

Any session by Kevlin is great, so this is definitely recommended watching: [Wayback/Archive] Structure and Interpretation of Test Cases • Kevlin Henney • GOTO 2022 – YouTube

Via [Wayback/Archive] Ifeora Okechukwu on Twitter: “On the structure and interpretation of test cases: by @KevlinHenney Awesome talk!!! “.

A few takeaways:

  • Kevlin uses the same Leap Year case as I have taught since the mid 1990’s: I like the way we both approach software development, though I can still learn a lot from Kevlin.
  • Non-programmers understand more than you might expect, just like you can get much more from a foreign news paper than you would think at first thought
  • You should know how to draw a Venn diagram with 4 sets allowing *all* combinations of those (search for “Venn’s four-set diagram using ellipses” in the link on the left)
  • Kevlin shows these 4 sets:
    • Common cases
    • Simple cases
    • Boundary cases
    • Edge cases

    The first two and last two sound similar. They aren’t.

  • Wrinkles and various levels of code coverage: don’t fool yourself as that is easy to do
  • The way of thinking is “thorough” (note from self: AI is not thorough, it is statistics)
  • Calendars are difficult (and depending a lot on the period and region their usage are valid for) – I already knew that, but it was good to be reminded off; see presentations by Jon Skeet for more information on it, see the blog tag Jon Skeet, or for instance these links:

Oh, Kevlin also covers FizzBuzz mentioning the enterprise edition which I already covered quite some time ago because Isotopp brought it to my attention: FizzBuzz as interview question – video by Tom Scott.

The reason Kevlin covered FizzBuzz is to again emphasise there is always a wrinkle, which you can see with his coverage hierarchy:

  • function coverage
    • statement coverage
      • branch coverage
        • path coverage

Even with all of these at 100%, you still can make your tests clearer: they are a fundamental part of your code base.

Query: [Wayback/Archive] jon skeet date time at DuckDuckGo

--jeroen


Leave a comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.