Testing is painful. There are always a hundred and one better things to do than test—but testing is essential if you’re going to write good software. So utPLSQL attempts to help coders overcome this natural aversion by creating a lightweight test framework that is easy to install, use, and run.

Any graduate of the corporate IT quality initiatives of the early 1990s will probably remember having this principle drilled into them by serious people in large suits. It’s a bit like English teachers throughout the world force-feeding their pupils Shakespeare: this sort of thing can turn people off for life. The utPLSQL project attempts to square the circle here by making test planning an essential and desired time-saving process, as opposed to something you’ll do when you have a little spare time.

This is an axiom of most open source developers; it’s a truism to say that the best coders are usually the best coders because they are the best testers. utPLSQL tries to encourage this approach, too.

In a bid to find the bugs and ease the workload, the utPLSQL methodology attempts to make testing as fine-grained and automatic as possible.

utPLSQL tries not to waste precious time studying complex test output. It tries to make things simple so you can know at a glance whether a test has passed or failed.

Don’t just fix code in a panic and on the fly. Build the alleged bug into the automated test suite. Provide yourself with both reliable evidence that the bug actually exists and reliable pointers to how it can be removed most effectively.