Hahaha, oh no you have also heard that "if statments" are code smell stuff now. It is just insanity.

Sure lots of if statements checking on type would be wrong, but haven't we known that for years. Saying if-statements are bad as a sort of general statement is so wrong on so many levels. God, how I hate this industry sometimes.

I am getting into more machine learning and data science now. Hopefully there people have enough challenges with algorithms to begin overengineering. Anyway it is Julia and I feel the Julia community isn't as polutted by overengineering as made other developer communities.

Julia suffers more from people writing too much "clever" code.

I actually watched Ian Coopers talk on NDC Tech town in Kongsberg a few months ago. I was there having a talk about RISC-V microprocessors.

I quite like his talk, but he kind of confirmed a lot of the views I have held about TDD for a long time. He talked about they buried themselves in test slowing down their progress. I had exactly the same experience which made me skeptical towards a lot of the TDD evangelism.

I am not against testing, but I am strongly opposed to the kind of "silver bullet" approach where we are supposed to rigidly follow some kind of awesome paradigm. I advocate REPL driven development: https://medium.com/codex/test-driven-vs-repl-driven-development-809d3c7a681

But I tend to like old well tested ideas. REPL driven development is extremely old. I have some affinity for old ideas which have survived. They must have some merit if they managed to survive until now.

Instead of the Uncle Bob, Kent Beck types of guys, I prefer listening to more of the old school guys such as Rob Pike, Joe Armstrong, Linux Thorvalds, John Carmack, Alan Kay, Brian Kernighan, Donald Knuth.

I think they have deeper and more profound ideas. They just are not as actively pushing them.