Now my first question to this is how to get to 40% when you are testing on multiple platforms?
For example, it takes 10 hours to test enhancement A on Windows. If that enhancement is required to be tested on AIX and Solaris, that estimate and effort are now 30 hours. I wonder if each platform should be a 1-off effort to make the 40%?? ideas? suggestions??