Friday, March 20, 2009

Software quality can only be understood backward

NOTE: the following is entirely hypothetical and has nothing to do with my employer

How many defects (or "bugs," to use the common but incorrect term) remain undiscovered by the QA folks, when a release is shipped to customers?

Well, you don't know; you can't know, until the customers find them. You can take guesses: last time the QA folks found X defects; Y were fixed, and the customers found Z more. This time, the QA folks found X' defects; Y' were fixed... but how big is the number Z' that the customers will find this time? We increased QA's budget, we introduced fewer new features, etc., so we think that Z' will be less than what we might otherwise predict from X' and Y', but....

But we have to make a decision on when we think the release ought to be shipped to customers before we know how many defects they'll bump into.

Thus, following the comments of that software guru Kierkegaard, we are forced to conclude that software quality can only be understood backward, but it must be lived forward.

No comments: