If it wasn’t for users...
Y2K analogies and mismatch
Being attracted by the wrapping
Scattergun versus scalpel
Testing the tools themselves
Welcome to the only presentation on the euro conversion at this conference!
I shall point up some challenges in testing and test automation, rather than focusing on the conversion process itself.
While I shall be describing some errors in existing software, the purpose is to bring attention back to well-established principles of boundary value analysis rather than to highlight specific failings.
1) The involvement of users in requirements specification and testing;
2) The problems with approaches that worked for Y2K testing and test automation but do not work so well for euro testing;
3) Two specific examples of testing the results of a popular technique known as “wrapping” or bridging or encapsulation ;
4) A specific example of how arbitrary test numbers used in an accreditation test missed an important rounding boundary condition ;
5) To begin with, a comment on the importance of testing the tools themselves: I supply euro calculators on some training courses and many participants just jump in and use them without checking that the correct conversion rate is being used.