Friday, April 20, 2012

STAREAST Day 2 Recap

I have to admit, it's a challenge to separate the fact that I am on East Coast time from the realities of waking up at 7:00 AM each morning, which is actually 4:00 AM where I'm from. I do that from time to time, but not with such late evenings. Still it's been a great deal of fun to hang out with such a great group of people and to have such wonderful, serendipitous moments.

The day started out with breakfast and a keynote address from Dorothy Graham of Grove Consulting about "What Managers Think They Know about Test Automation–But Don't". Dorothy points out that often, Management is looking for the following aspects to happen when they talk about test automation. They believe:
  • automation can test everything
  • automation will find all the bugs
  • with automation we will get test coverage of 100%
  • more automation means more confidence
  • run tests overnight and weekends
  • automation will prevent human error
The fact is, there's a lot more that goes on in these areas than what management hopes to have happen. Yes, automation can save time, but it will only save time after the automation is developed, and it won't save time if the development of the automation is close to the amount of time it takes to test things manually. Trust me on this, it takes time to automate tests well. The architecture of well-running automated tests is key. making that architecture takes time. Poor architecture often leads to poor automated testing, and frequently to abandoned automated testing. The relationship with the development team is also critical. If we work with them, we can develop solid tests that can be useful for the long term. If we develop our tests in isolation, we may miss a lot of things, but more importantly, we may find we are working at cross purposes and not realize it.


James Bach gave an excellent talk on "The Dirty Secret of Formal Testing". What made this talk interesting was the fact that he was describing testing in highly regulated markets, dealing with a product that was very life or death (a medical device to help aid patients with heart problems). These environments are extremely formal, but the dirty secret of these environments is that informal testing goes on all the time. It has to. Otherwise there's no way to learn about what the product actually does. As described by James, formal testing is any testing that must be done a specific way or must check specific facts. Requirements as described by James (and I like this description) are ideas at the intersection between what we want, what we can have, and how we MIGHT decide that we got it. In complex and life-or-death systems, we have to test to find out if that has happened, and we have to test first to find out if we actually accomplished those goals. Thus, contrary to popular belief and conventional wisdom, good formal testing MUST first begin with informal testing.

In addition to the variety of sessions, I have to point out that many times, the best discussions and best ideas happen between the sessions, or happen when we decide to engage in conversations with a single person or a handful of people, effectively creating our own session on the spot. I confess, I had a hidden motive with coming to STAREAST beyond giving my talk and learning some cool new ideas to bring home. My goal was to see what I could learn from many different people about how to teach software testing, and make it fun, for 16-24 year olds. More to the point, I needed to recruit helpers willing to do the work, and I was not disappointed. Instead of me looking for people to discuss this idea, word preceded me, so I had people coming up to me to ask if they could take part. Based on the interactions I had with many people this week, I got what I came for and plenty more. I had some terrific discussions with people from many industries as well as educators who have helped me narrow down and distill some of the ideas I have, and it's given me many new directions to consider. They may not all be possible, but I'm looking forward to trying them out.

STAREAST Virtual was a fun and interesting way to communicate with the audience of participants that were not specifically at the conference, and it was fun to have Matt Barcomb participate with me in a conversation about Testing Roles and Approaches to Paired Development and Testing. In addition, we also discussed many of the aspects of collaboration and ways to break down the unintentional walls that we build. In this way, we can set up our environments and approaches so that we can become more effective.

The main conference came to a close with the awarding of a number of awards, including those who found interesting bugs in the Test Lab, and something that made me smile greatly. It was announced to everyone just before the closing keynote that my paper and presentation were chosen as Best Paper for the conference :). For someone who gave his first official full hour-long track talk at any conference, well, yesterday, that was a wonderful thing to see happen (I rank it up there highly with "firsts" of my career :) ). I appreciate the award and those who felt I deserved it. Thank you very much!

The final keynote of the conference was given by Theresa Lanowitz of voke, inc., and it discussed "Testing Trends: Cloud, Virtualization, and Mobility". It gave a forecast of what we might see in the next two to three years, and I must say, things look exciting for testers, regardless of what the "test is dead" people are saying :). Numerous examples of performance problems, unintended consequences, security breaches and other issues of integration were explored, and they all pointed to the same thing. Testing needs to be performed, and it needs to be performed at a level that people can make solid and intelligent decisions about the ramifications and consequences. The simple fact is that there are no end of authentic problems to be tackled and the scope and the landscape are expanding. There will come a time when we will need more testers, not less. Get ready, folks :).

My thanks to so many people for the time they took to get to know me, speak with me, hang out in the hotel lounge or go out to dinner with me and talk about ideas and concepts that we all deal with and the solutions we hope to achieve. My thanks to Scott Barber, Claire Moss, David Gilbert, Matt Barcomb, Lee Copeland, James and Jon Bach, Lanette Creamer, Rachelle Sawal, Janet Gregory, Allison Wade, Matt Barcomb,  Randy Rice, Mirkaya Capellan, James Lyndsay, Michael Bolton, Zeger Van Hese, Griffin Jones, all the attendees of my talk about Weekend Testing, and everyone else that helped make this a memorable week.

'Til we meet again :).
Post a Comment