The final SIGiST (Special Interest Group in Software Testing) conference took place yesterday in London and, as usual, was well worth attending. The theme for the day was “Keynotes – Six of the best” and consisted of talks only on this occasion: six keynotes and one short talk after lunch. Unlike other SIGiST conferences I have been to there were no workshops which I always enjoy but I still found the day inspiring.
Four talks stood out for me as being excellent:
Les Hatton from Kingston University gave a brilliant talk in which he cited the systems controlling space shuttles as an example of excellently engineered systems and then went on to talk about systems which “should never have been allowed to see the light of day”.
One of the ‘highlights’ (that should probably read ‘lowlights’) included the story of his passage through Heathrow Airport earlier this year. He had printed his boarding card at self-check-in but the systems at security could not read the card; SAS (the airline he travelled with) could not issue a new boarding card because he already had one unless he gave them the assurance he was who he said he was (!); he was then unable to get through security because he had two boarding cards… As if it could not get any worse the public information displays in the departure lounge had crashed. As a keen traveller I could really identify with Les’ frustrations here!
The final part of Les’ talk encouraged us to focus on systems thinking and take some of our cues from the laws of physics. Once you find a bug in a particular area of the system you are likely to find more bugs in that same area. Don’t give up was the message.
Gojko Adzic gave his excellent talk on Specification by Example. Once again he made very good use of Prezi and encouraged us to use clear and concise language that our colleagues and customers will understand. Too much time is wasted by misused terminology. In a later talk mention was made of test cases and test conditions – actually they could both have been referring to the same thing – so why distinguish between them?
As usual Gojko had lots of examples to illustrate the success of the technique. I find the concept of ‘living documentation’ particularly valuable and I liked the example of customer service staff referring to the tests that had been run to help answer customer queries. It makes the tests very powerful because each test is directly addressing a particular problem being faced.
In the afternoon Fran O’Hara from Sogeti Ireland gave a talk on Scrum. Included in the delegate pack for the conference was a Scrum cheat sheet illustrating the different components of a Scrum project and showing how they fit together. I took it into work today and our Project Manager has found it very helpful. I thoroughly enjoyed Fran’s talk and particularly liked the idea of having two definitions of ‘done’: one definition describes what it means to be ‘potentially shippable’ and the other defines what ‘done’ means in terms of the current sprint. There are many projects where it is not feasible to produce a potentially shippable product after one or two three-week cycles and this helps to deal with that.
Susan Windsor from Gerrard Consulting finished the day talking about how we develop ourselves and what it means to be a really good tester. Susan challenged each of us to become testing gurus, super-testers in our organisations. This will pay dividends because of the tremendous knowledge that we can bring to the table of how our projects are really going whether we are working in a traditional or more agile context.
Susan discussed the certification issue briefly and reminded us that although we can get a sense of achievement out of having a certificate one of the biggest problems with certification is the fact that it is used as a screening mechanism by people who really know very little (if anything) about testing when hiring staff. Personally, I would add that the syllabus is too restricted in its scope and is based on very traditional testing processes which have been shown to be less efficient than the agile methods being adopted more and more.
Other talks included a career progression report from Erkki Poyhonen where he experienced a paradigm shift without a clutch (cue a Dilbert cartoon), a report of an entity-relationship modelling exercise for testing effectiveness from John Kent and a short talk from Geoff Thompson on the things that have influenced him in testing.
As always the day ended in the Volunteer where I enjoyed continuing chatting to fellow testers about our exciting craft. The best thing for me about SIGiST is the networking and getting to know other testers. As a result of attending these and other conferences I have built up a network of people with whom I communicate regularly and it has really expanded by knowledge of my chosen craft. I would encourage everyone to get involved with testing conferences in their various locations because together we can learn a tremendous amount.