Back by popular demand, James Bach has returned for another Testing the Limits interview. An author, speaker, coach and consultant, James is one of the testing industry’s most compelling figures – constantly challenging conventional wisdom with style that is his alone. If you’re new to James Bach, you should – nay, must – spend a day or two on his blog.
In part I of our two-part interview, James discusses “pathetic compliance” and other newly coined terms; weeding out bad ideas in the field of testing; the increasing popularity of context driven testing; Apple’s antenna bugs and more. Be sure to check back tomorrow for part II.
uTest: When we interviewed you last year around this time (here and here), you had just published Secrets of a Buccaneer Scholar. What have you been up to since then? Any new books in the works? New speaking opportunities? New theories? New hobbies & interests? Said differently, what’s coming in the next 12-18 months for the Buccaneer Tester?
JB: I coined the terms “test strategy playbook”, “pathetic compliance” and “thread-based test management.” I’m working on a book with Anne-Marie Charrett on how to train testers over instant messaging, and done a lot of online coaching along the way (I offer this for free on an as available basis, or if I’m paid I’ll schedule a lesson). I joined the board of the Association for Software Testing, again.
The main thing I’ve been doing is being test architect for a class III medical device. I got to read lots of standards and regulations, design a statistically justifiable accuracy testing protocol, create a log file analyzer with automatic graphing, document a mathematical algorithm, and I developed a tool for reshuffling pages in dozens of PDFs to create topical doc repositories.
uTest: Your domain name is satisfice.com – what does the “Satisfice” name mean to you, and why did you choose this name?
JB: Satisfice is a word coined by Herbert Simon, who won the Nobel prize in Economics (technically, it’s the Nobel memorial prize or something like that) for his work on bounded rationality. Satisfice is a verb meaning “to achieve a result that is good enough.”
Much of what I do and say is based on the work of Herbert Simon. He’s one of my heroes. Also, the domain name was available…
uTest: Is the testing craft better or worse off since that time (Nov, 2009)? Which developments or trends point us in either direction?
JB: I don’t know if the craft has changed in any significant way, but my business has picked up a lot. It must be the economy getting a bit better. By the end of the year I will have taught in ten countries in the last 12 months.
uTest: You recently said that, “There is no such thing as an ‘exploratory tester’ except inasmuch as a good tester obviously can and will do exploration as a basic part of his work.” Are there too many “labels” in testing? If so, what are the consequences?
JB: I would say we need more and better labels, while retiring some old ones that don’t help. The consequences I’m worried about are what happens when people say they are tired of this profusion of terminology. That’s like saying “I’m tired of software companies creating new technology! Let’s go back to HTML 1.0!” That’s like saying “I’m tired of learning. It hurts my brain. Everyone has to think the same thoughts now!”
I’m sorry that testing is complicated, folks. No wait. I’m not sorry at all. Go away if you don’t like it. I think we will be creating new ideas and labels at a good clip for years to come. Let’s do a good job of it, though, and weed out the silly ones. (For instance, no one should utter the term “boundary value analysis” ever again, unless they are talking about actual analysis, rather than the simple act of adding 1 and subtracting 1, which is a form of “analysis” any six-year old can routinely do.)
uTest: The amount of testing-related content on the web is exploding. At the same time, the Context-Driven school of Testing has never been more popular. Coincidence?
JB: I wouldn’t be surprised if CDT has never been more popular, because we produce lots of new ideas, actively weed out the ones that don’t work so well, and we share them freely. Go to any practitioner’s conference and you’ll either see the Context-Driven community frequently referenced, or else you’re at an ISTQB/ISEB Zombie Rally, (or whatever they call their dark gatherings by moonlight when they invoke whatever evil spirits and marketing consultants that help them deceive the unwary.)
Also, the Association for Software Testing has officially declared itself a Context-Driven Testing organization, and the first conference dedicated to Context-Driven Testing will be held next July in Seattle (the 2011 CAST conference, which will be run by my brother and I).
Context-Driven Testing represents the free world of testing. We’re sharper, braver, hipper, and we believe in working at our skills instead of smugly assuming that we are already skilled enough, so of course it should be popular… Except…
Except three things:
- I don’t actually know that it is more popular then it ever has been.
- If it isn’t more popular, there’s only one other possibility: that’s it in decline. That would make me sad, so I hope that’s not true.
- The second point was such a downer I can’t think what else to add to this list, but there’s probably other caveats.
uTest: The divide between hardware and software seems to be narrowing. We’re thinking specifically about the problems encountered by Apple (antenna) and Toyota (brakes). If true, how does this impact the craft of testing? Is there a role for software testers on the hardware side of things, and vice-versa?
JB: Oh yes. The Way of the Tester applies to anything.
(Caution: Don’t apply it to your wife and kids, unless you like grumpy housemates.)
It’s all just rapid learning, analysis, questioning, the relentless exploration in search of order and disorder. I’m not an electrical engineer, but I had to learn about electronics for my recent project. Give or take an Ohm’s law and a phase angle or two, it’s all the same stuff. Testing is testing.
uTest: Along those same lines, serious bugs have proven to be as much a PR problem as a development problem. Any advice for companies that must deal with their bugs in the public spotlight?
JB: Yes! Come clean and stop complaining, Steve Jobs! Of course your products have problems. And once in a while they are serious. Instead of whinging about it, have your PR firm get the test manager out front and having him talk about the challenges of producing excellent products and how the best companies are distinguished by how they learn, adapt and respond, not by how they refuse to take risks or admit mistakes.
You can make people trust you not just be never making a mistake, but also be showing that you can make good on your service despite mistakes.
Editor’s note: Read part II of the interview.