ACCU 2010 - Exploratory Testing with James Bach

Wow....where do I start?....

Well I attended ACCU 2010 at the Barcelo in Oxford Yesterday..


I had a mixed fortunes 'early morning' - waking up with a dodgy tummy....but followed by a black cat walking across the road right in front of me on the car journey to the venue, I took this to be a good sign and soon forgot the dodgy tummy...

I attended the session with James Bach entitled 'Exploratory Testing Explained'. I have to say it was awesome and there are some things I can instantly use from the session....

As far I remember James started the session talking about self defence in testing....what? you may ask....

ok....so have you ever had anyone come up to you and ask you whether your doing something you've never heard of.... i.e. Are you doing 'super douper equiv part version 1.2.1'? Have you done 'blahdeblahblah'?



Rather than a 'no' response it would better to respond by using a phrase like 'Not known by that name' but we're doing this that and the other...

My own thoughts on this also included asking the person what they mean by the phrases they're are using...but I guess by responding like that may take time, while a 'Not known by that name' response gives instant feedback to the questioner


James stated that exploratory testing is an approach and not a technique, exploratory testing can be applied to techniques...


We moved onto the distinction between scripted and exploratory testing




Thinking of scripted testing as a straight line

and exploratory testing as loops.

James talked about the idea of 'soft scripted testing'....and who is generating the test ideas, are the test ideas under the testers control...I think there probably should be cartoon for this...where's the cartoon tester?

Exploratory testing can have scripted elements for example using a tablewhen testing (scripted) but changing the data within it (sorry for the bad explanation) I have done this before and have been meaning to blog about this for a while...A good prompt from the session to get that post out.

We did several exercises, firstly where James presented us with an application to generate the minimum number of tests...

A few tests later.....

In fact there were no minimum number of tests which James called 'epistemic infinity' - I think this means something finite with no upper bounds....

We did another exercise with dice, we could throw dice(of all shapes and colours) say a number and James would then say a number back to us. We had to work out James algorithm. I learned a lot out of this exercise, because it was all out about focusing and defocusing....

I would like to have done a lot better at this exercise, however because I didn't do as well as I'd liked it's kinda etched on my brain.....which can only be good right?

I think I was probably thinking about lots of things and tests in my head all at once.....i.e I want to try this this and this rather than focusing on one thing at a time. Or I should have fused these strategies?...interchanging focus and de-focus...

I did contemplate asking other people about what they were finding to get a better spread of data in order to perhaps get to the algorithm quicker....but would that have confused me even more?

James mentioned the Car Park Calculator to everyone (I think this causes a twitter storm when Matt Heusser presented the challenge) I briefly had tested this before when Matt originally released the challange......How did you fair with that challenge? What's is the highest amount that you can get? Good practice?

James demonstrated blink testing....

Mentioned that you can't rely on your memory for recording exploratory testing. James uses video to record sessions...(this is quite interesting as I have recently done some videoing of my own). I also have previously used tools like Wink and fiddler. James also advocated function level logging. I think this is great....there is lots of logging going on with the system I'm currently testing...however, I have one caveat to this...You don't want to spend too much time testing the logging!!....

James mentioned visualization and a tool called ( NodeXL) - which I may look at using amongst other visualization tools - look out for future posts...


We also covered sessions.

The whole day was great, lots of vareity interaction, magic tricks and it most definately kept my interest throughout the entire day...'there was definately no afternoon switch off'. At the end of the session I felt more energized...

The things I will instantly do:

1. Blink testing......I currently working on a system that has huge log files and I have been searching the log for specific things or blinking at specific searches but I think now I'm just going to expand that and look at the whole log and see if my brain notices any patterns..

2. Focus and de-focus - Focus meaning small change, de-focus meaning lots of change
I need to practise this a lot.

3. James taught me a magic trick which I'm going to try and perfect through practice...

4. Practice, Practice, Practice and more practice.... ( I quickly realized that I need to do a lot more - I may get to get coaching sessions over skype yet!)

Is there a theme here?

Practice? What are you practicing?