Brief

A brief post today, as I’ve been working rather long hours. I have things on my list that need to be no longer on it - and as of now, I think I’ve knocked them off. Phew!

Best win of the day: I hope I’ve solved a “flaky test” issue. In programming as in real life, tests are a useful trick to find out where things stand. A good test is reliable: it tells you what’s what. In software, and indeed other areas like healthcare, a test tells you whether everything works as expected or not.

So what happens if your test isn’t reliable? If the program works fine, but the test says it isn’t doing the right thing? Or the other way round - it’s broken, but the test says all is good? It’s as much as a problem in software as it is anywhere else: you need to know where you stand.

A flaky test is one that doesn’t really tell you whether things are working or not. And today I figured out one particular source of this uncertainty and - hopefully - fixed it. Tomorrow I’ll know whether this was successful.

Photo by Pankaj Patel on Unsplash