The One with the T Shirts

This is the “mockery of a presentation” I gave at the Israeli latest .Net user group. Now, apart from the pure genius that the presentation includes, I used a couple of tricks in it.

I wanted to the audience to remember to things. The first one is the number (41%) which appears in slide #16. For those who weren’t there, that’s the number of extra code lines for hand rolled mocks, which I banged on for a couple of minutes – it symbolizes the same amount of time wasted by developers who use hand rolled mocks instead of using a framework.

The other thing I wanted them to remember is ME. I hope I’ve managed that by using humor, and a bunch of T shirts (see presentation again). I have a T shirt appearing in almost every slide, which somehow relate to the slide content. And to be honest, how the T shirt relates to the slide is not important. What is important is that they appear in EVERY slide. People were actually looking forward to the next slide to see “what is he going to show next”? (I was told). And I’m sure I’ll be remembered by some people as the “one with the T shirts”.

For all tech geeks out there: You may argue, that I should want to be remembered for my unit testing expertise. I do as well. But that’s not how people, even tech geeks, remember. What we remember is the difference, the change. Sorry, but that’s how our mind works. And so using “tricks” like these are necessary to leave an impression.

And so, at least for now, I’ll be the T shirt guy.

2 comments on “The One with the T Shirts”

  1. Aviv Ben-Yosef Reply

    Nice presentation. I went through the slides, and saw no mention of how you got that figure (41%).

    Also, have you seen uncle bob’s blog about overusing mock frameworks ? link

    It might the the case with C#, Java and other statically typed languages, though I wonder whether rolling your own mocks in dynamic languages will be different?

  2. Gil Zilberfeld Reply

    Hi Aviv,

    Thanks! The 41% was calculated from the difference in code lines from the generated tests (with hand rolled mocks) and the Isolator framework test. The number of course is an example, but the range of 30-50% is the majority.

    Uncle Bob exposes the magic for the benefit of readability. I disagree with his example which is fairly simple. In real life you’ll have much more complicated mocks, and therefore a nightmarish maintainance issue.

    Dynamic languages are a whole other animal. mocking in Ruby is much more simple – you just change the method – no magic, no framework. It’s part of the language. But you mess things there as well, if you’re not careful.


Leave A Reply

Your email address will not be published. Required fields are marked *