Never Look Back
I gave a keynote at CUSEC in 2010 about evidence-based software engineering. A recording was available on Vimeo for several years, but it disappeared some time after 2016. A few days ago, though, I got mail out of the blue from someone who had found a copy. They shared it with me in case I wanted to post it somewhere, so I downloaded it and watched it through, and now I feel a little bit sad. In increasing order:
-
A lot of what I said was wrong, either because the claims I made were stronger than they should have been or because the science itself has turned out to be flawed. (Stereotype threat, anyone?)
-
I was rude to a couple of people in order to get laughs from the audience. I don’t expect any of them will read this, but if they do, I’m sorry.
-
I really did believe things would get better. They haven’t: students don’t learn any more about research practices or empirical results in software engineering today than they did fifteen years ago, and programmers in industry are just as oblivious as they ever were. I really did think that Making Software and It Will Never Work in Theory would have an impact on more than a few scattered individuals; I’m disappointed that they didn’t.
I’ll post the video in a few days, but I honestly found parts of it hard to watch. I’ll be speaking at CUSEC again in January; here’s hoping that talk holds up better.
Time to make another cup of tea. If you came in peace, be welcome.