on 30 December 2005
This volume is rather prescriptive, giving the LogicaCMG approach to test management. However, there is something in the volume for all, and you do not have to be part of the circle around this firm to get something from it. The proviso is that if you are NOT part of the circle around this firm, you should not possess this as the ONLY book on test management.
I found the introductory chapter rather simplistic, describing the historical testing methodologies that there have been. All of these were found to be deficient, when, over the horizon gallops John Wayne, and the methodology known as Risk and requirements based testing (RRBT). That may well have been true, but “risk” has been a well recognised part of the mainstream testing for a number of years. Therefore, the justification for using RRBT is almost taken as read by the majority of those looking into the volume.
Having said that, there are items that give a different insight in the pages written. This reader found the ‘quick scan’ undertaken at the start of test planning an appealing concept. This stage is undertaken in order to build up what the authors refer to as the ‘Test management File’. This ‘file’ was not well introduced, and it was some time before it was realised that this is not so much a file, but a method of organising information. A meta-file would be a better way of understanding this. Perhaps something has been lost in the translation from the Dutch.
The chapter on test estimation was particularly helpful, with good sections on why estimates are often inaccurate, what items are often left out of estimates (including ‘test management’, would you believe), and what to do if the estimate, however it is achieved, is not acceptable to the stake-holders or others who are footing the bill. At this stage, if the estimates are too high, there are very stark choices; either risk are addressed, or they are not.
Key elements of RRBT is that testing can stop at any point, and if that were to happen, the BEST tests would already have been performed, giving the best value to the business. Four test types are recognised by the authors, although these do not necessarily relate to test cycles. These test types are the intake test, the basic test, the complete test, and the final test. There is an order of dependency within these; if the intake test is not successful, there is little or no point proceeding.
The appendices are for the most part an excellent source of additional material. Of special note is the roles and responsibilities of those involved in testing. This would be useful in the staff selection process. One point where the supplementary material did break down was in the glossary. This is a little surprising as Erik van Veenendaal is one of the authors, with extensive experience of glossaries. It could be it was felt that a comprehensive glossary was unnecessary – perhaps this should have be stated more clearly (I did not find this sentiment referred to).
The final chapter discusses the transfer activities, and discusses where lessons can and should be learnt from one testing project, to be carried on to other future testing situations. There are other sources of material available on this extensive topic; having said that, the coverage was limited.
So, would I recommend the book to others? My big reservation to those outside the LogicsCMG sphere of influence is that I believe that there is no set answer, no ‘one size fits all’ approach. Thus the answer is “Yes, ……………… BUT”. Have other sources of material, and question the advice given. Take what is good, and useful to you, and leave other elements. Use the volume to help you answer why you do what you do, by taking a look at something else (i.e. the LogicaCMG approach). It should be noted that some other volumes on ‘Test Management’ also cover, for example, test techniques, which are not within the scope of the present volume.
If you are within the LogicaCMG circle, the decision clearer – buy!