June 27, 2014

Sometimes you just need a big red button

When clearing my office yesterday, I came across a dust-covered gadget that instantly brought a smile to my face. In one of my previous roles, it had proven to be the defining part of introducing an infrastructure for automation.

It was a big red button.

Here's the scene: I was introducing an automation framework from scratch, and I thought I'd already done the hardest part. I'd built a suite of test-runner VMs on our ESX server, and created scripts to deploy the build-under-test in each environment. I'd built a well-encapsulated GUI smoke test in QTP, with reusable library functions to allow more checks to be added easily. And I'd built a reporting process which would collate the results of the test run, and email it to the test team.

Yet I was still struggling to get buy-in from many of my QA colleagues, who weren't showing an interest in adding automated checks for the functionality that they were testing. Similarly, I think the development team viewed our automation process as "something which Neil runs from time-to-time for some reason", without being particularly curious to find out more.

The process itself seemed to be working fine, but promoting it was proving problematic. So I looked at what I could change, and settled on a USB Panic Button (£1.99 on eBay):

The button

By default, the software would only spawn a full-screen JPEG when pressed, mirroring the "boss key" of old. However, with some homebrew drivers that I found online, I was able to reconfigure it to launch our test runner.

Naturally, curiosity got the better of people, and in the weeks that followed, colleagues from all disciplines and levels were asking me "What does the button do? Can I try it?"

My fellow testers even vied to be the person who launched the tests when required, which (over time) gave them the knowledge and confidence to extend our scripts.

In essence, for less than the price of a fancy coffee, I'd added a tactile element to our testing framework, which completely changed how our work was perceived.

As the months passed, we didn't even need the button any more. It had served its purpose, and as the team extended the framework, we introduced a continuous integration process, decreasing the need for ad-hoc test runs. The button was retired to my drawer, but its legacy was secured.

It felt a little embarrassing to write this post, because (as a technical tester) I'm really much more interested in discussing the automation harness itself: what worked, what could be improved, what I would change next time. But in this particular scenario, and as Father Dougal discovered in Father Ted, there's something so darned irresistable about that big red button.

Do not press

And no blog post about a big red button would be complete without sharing this classic video from James and Jon Bach, where they demonstrate a range of exploratory testing techniques during their investigation of an "Easy Button" from Staples.

  • LinkedIn
  • Tumblr
  • Reddit
  • Google+
  • Pinterest
  • Pocket
Comments powered by Disqus