We are going to need to produce documentation about the efforts that we're undertaking in our testing process. Because what we document, how we perform it and the results that we get, are of very great importance to our future efforts. So, we document what has been tested, what was found or what was not found, what the cause and effect of all the different things that we've tried were, and how, if it was, fixed.

Now, relative bug priorities are like anything else we encounter in a process like this. Not everything will have the same priority, and we can only have one 'number one' priority. But once we establish these, we also need to establish what action was taken with respect to the type of priority or the level of the priority. Obviously, we have to have levels because we need to have things put in an order, most important first, most damaging first, working our way down to least.

We're going to have to have contingencies or fallbacks to ensure that we have different routes to get to the solution, whatever that might be. And these need to be described because there will not be just one way to fix something. And we need to describe what those ways are and why we chose the one we did. Each scenario thus must be memorialized in our documentation to facilitate the necessary CMMI-type process improvement. It means, of course, we're going to have to have careful version control throughout, so that we can ensure that whatever was found and fixed, is not, at some future point, returned to a live state, and thus, once again becomes a problem.