_Every_ avenue of risk, even the ones that they don’t yet know about.
I am learning that the subject of testing parallels religion. There are those that believe and those that don’t. At the end of the day, it comes down to you either believe in God or you don’t, for no particular reason, you just do or do not. Sometimes people are in the middle and can be persuaded one way or the other.
Software testing is kind of like this. Some believe, some don’t, and some can be persuaded one way or the other. The interesting thing is that in regards to religion, those that believe just do, and can’t be convinced otherwise, no matter what sort of empirical or rational argument is provided.
Testing evangelists find similar situations when proselytizing testing. However, they often only use arguments from reason, or a priori arguments as they are called in epistemology. The other kind of argument, namely, the argument from experience, in epistemology is called a posteriori. I suspect that testing evangelists will have much more success using a posteriori style arguments. Rather than hoping that non-believers suddenly see the light and start testing, perhaps the evangelist could walk the non-believer through the ropes, letting them experience all the benefits of test driven development for themselves. It may not be enough to win new convert, but it might be enough to make the a priori arguments of testing that much more effective.
I love being a hacker. It’s like playing with legos, except that you can build any sort of app that you can imagine. This morning I was given a simple problem, ‘how many open/closed pull requests do we currently have?’ Within a couple hours I had built a simple one page app that would give you the answer.