I will start that I think TDD is very valuable. But reading this article that quickly goes into a numbered list of "rules" is a little disheartening. Already so

I will start that I think TDD is very valuable. But reading this article that quickly goes into a numbered list of "rules" is a little disheartening. Already some other commenters are talking about more nuanced views of when to test. I think the real tragedy is that I see a lot of engineers still not writing tests. So when speaking to the broader audience, you have to be really gung-ho about it to get people to do what I think is a very necessary step. People should err on the side of writing the test first, both to get most people in the right mindset, and also to avoid the broken window syndrome of a lot of missing tests and no one bothering to actually start doing what should be obvious: most code should have at least some level of coverage. Most in fact. If there are "reasons", sure, if you're going to pull the senior engineer move and skip the test, then you better be able to back it up clearly. If not, write the damn test. If not you, someone will appreciate it later.