I've developed an add-in for Microsoft Excel 2007 and validated it by covering all the software requirements by tests, which have been executed, recorded, analysed and documented.
I've made a new version of the add-in for Microsoft Excel 2010, the functionalities are exactly the same (it's the same code), but I've just changed the setup project to create different registry keys during the installation, and the version of the .NET framework used is different (.NET 4 instead of .NET 2).
I'm not sure about what to do for validating this new version of the add-in in a quick but correct manner. As the code has not changed, I think I should not re-test all the software requirements.
I'm thinking about documenting a risk analysis to list the changes between Office 2007 and Office 2010 that could affect the add-in and determine if some functionalities have to be re-tested. Then testing these functionalities (if any) and the new installer (by checking that the add-in and the pre-requisites are correctly installed, and the add-in is correctly loaded in Excel after installation), and make some general testing to verify that the critical functionalities are working.
What do you think about this approach, would you have some comments or advice?
Thanks in advance,