testjutsu

In testing, I think there is a tendency to confuse repetition for the purpose of learning with repetition as the application of skill. I have heard that some expound using scripts as a means of skilling up unskilled staff. I believe this thinking is fundamentally flawed. The reason being is that there is no clear delineation between the repetition of an action (i.e., a script) and actual software testing, which is far, far more than simply following a recipe.

In kendo, a beginner starts by learning the basic moves. A student might practise a single cut 10,000 times or more. Time and again they are made to demonstrate correct footwork, body position, striking and movement. At this point, they have no clue about the application of these techniques, only how to execute them. They can make a cut, but it doesn’t mean they can consider themselves skilled in kendo.

Once they have the basic mechanics down, they are ready to strap on armour and face a real opponent. Unlike before, where they had specific instructions about what do do, how and when, they now face a situation where the application of their techniques may or may not work and it is up to them to work out why. Moreover, their opponent is attacking them, shouting at them, trying to unsettle them, make them nervous and throw them off their game. Are you software testers out there seeing parallels?

The difference here is that the delineation in kendo between learning basic technique and application of that technique is quite explicit. In software testing, especially where scripting is involved, this is not always so. Herein lies the danger of using scripts as learning tools. A script may help you learn to execute a technique. It will not teach you to be a tester.

So I’m back from CAST. Yes, I know I haven’t finished writing up my notes from the CITCON event. I had every intention of doing so on the trip home, but 30 hours stuck in planes and transit lounges with screaming children, people with questionable personable hygiene and vapid narcissists who need to share with everyone within earshot is not conducive to cogent writing. The good news is that I’m not in prison for visiting bloody ruin upon any of the aforementioned human vermin, and hope to return you to your (ir)regular programming soon.

So, CAST rocked. Toronto was a cool place. The local people were, if not friendly then at least not aggressively disinterested (apart from a bartender at the Imperial, who was a total douche). The attendees of the conference were amazing human beings, if for no other reason than their eyes not glazing over when I was talking about testing. Some really bloody smart people there doing some very cool stuff.

I took the day-long tutorial with Jerry Weinberg on communicating as a tester. The cost of the course was probably worth it for this alone. There were keynote speeches by other big names in testing, which were also very good. More importantly though I think were the people I got to hang out with. I think it takes a particular brand of weirdness to be a good tester. There’s a certain pattern to the way they speak that gives you an insight into how they think. Testers aren’t afraid to probe, or question either so I also enjoyed the distinct lack of listening along politely.

I have more material to write up than I know what to do with. I have a busy few weeks coming up on the kendo front, so I’m not going to get to it as soon as I want to, but it’s coming. In the meantime, check out the CAST site conference archive. If you’re considering going to next year’s, it’s in Colorado Springs. I highly recommend it. I shall be doing everything in my power to make it there. If you have any sort of interest in software testing, I urge you to do the same.