Five Programming Books That Changed How I Think

Obviously, this is a timeless classic. I too love the writing style and its precise, articulate prose; it sets a bar for what I expect at MIT. There is much wisdom to glean.

This note about "wishful thinking" is an important concept from the book that I rarely see mentioned, but is core to the authors' layered approach of building systems. In essence, it's a development style of hallucinating or confabulating appropriate functions and data structures during the design of your programs from the inside out, which you would implement later on. (A quite natural approach with a Lisp/Scheme REPL.)

It's also why I never understood the fuss about supposed failures in AI code generation. Some developers never read SICP, I guess.

> Paradigms of Artificial Intelligence by Peter Norvig. Learning the AI described in this book probably won’t land you a job today. But reading the code examples will transform how you think about source code. The book shines when it comes to code comments, a topic that I’ve never seen demonstrated well in other sources. Here we get to see how comments become valuable as a narrative that explains both intent and reasoning. Brilliant, just brilliant.

First, that's not the title of the book. It's Paradigms of Artificial Intelligence Programming. Second, why would you recommend a book because of comments in code. I checked the code, it has good docstrings, it does not have that many comments beyond that. The commentary (rather than code comments) is the book itself.

And I like this book, but the reason to recommend it is not the code comments (which barely exist outside of his very consistent use of docstrings). The reason to recommend it now (versus back in the day when GOFAI was just AI and the book had more obvious utility) is that you get to see how an expert programmer thinks, Norvig is among the best expositors of programming and CS topics, and seeing how he crafts programs to solve interesting problems is insightful. What's also nice is that he often doesn't start with a worked solution, but also shows some false starts and how he would progress from conception of an idea to trying some things out to a good solution. You actually get a glimpse into how he thinks through things.

So yeah, good suggestion and it impacted me (had it for a course in grad school decades ago), but this "review" is awful and is entirely unconvincing as a recommendation.

The rest of the reviews are only slightly better.

Rod Johnson showed how you could code in a lightweight, highly productive fashion with Pojos (Plain old Java Objects) and still scale. The principles were generic but became the foundations of the Spring framework that took over the ecosystem.

"Hibernate in Action" by Gavin King is another that came out around the same time. Again, a book on generic ORM principles, focused on a particular library. It also rejected the old orthodoxy and became the new standard.

——-

And I was recently thinking about books that my opinion has significantly changed about the longer I’m in my career, I realized that “Peopleware” truly sucks. It is such feel good fairy tale, that actually harms the reader. The real world is much different and much more cynical.