Ask HN: What are the most eye-opening textbooks you have ever read?

- CS:APP (Computer Systems: A Programmer's Perspective), as someone who didn't go to college, this made me feel like I knew the systems inside out and it was no longer magic.

- SICP (Structure and Interpretation of Computer Programs), this one is controversial, but getting to the end and actually building the interpreter made me see that seemingly complex programs like compilers and interpreters are just other programs and quite similar to the ones that I was used to writing. The other effect of this on me was learning about powerful abstractions. It felt smooth to follow along and build the interpreter because the abstractions in the book followed a logical progression. I want the other complex programs I write to seem just as simple.

- Software Foundations: Logical Foundations. Writing programs as proofs was fascinating. It felt like mathematical mumbo jumbo before but actually trying it out removed the magic and it became more obvious how programs and formal proofs could go hand-in-hand. The impact of this book wasn't as generic as the other two. Maybe one benefit of seeing programs as proofs is that by trying to write programs that can be proved easily, we write programs that are actually simple to reason about and explain to someone else.

1) Sakurai and Dirac's textbooks on quantum mechanics.

2) Axler's book on linear algebra. That's when I started to think in terms of vectors as more than the typical "magnitude and direction" concept you are used to in UG physics. It took many years, but I'm finally beginning to apply this to more advanced topics.

3) Concurrency in Action, by Anthony Williams. The training in multithreading aside, that motivated a deeper interest in things like how caches worked or atomics for someone without a CS background like myself. I haven't finished Three Easy Pieces yet, to my eternal shame as I read this thread, but that might also be added on this list by the time I'm complete.

4) On the right-hemisphere side of the brain, books such as Christopher Clark's "Sleepwalkers", Jonathan Smele's treatment of the Russian Civil War, and Richard Thornton's "Odd Man Out" on the Korean War have been useful in teaching me to dig deeper than the standard narrative when looking at history. I'm looking forward to Christopher Goscha's book on Dien Bien Phu and James Howard-Johnston's recent drop on the great final war of classical antiquity in the 7th Century, in a similar vein: both are out, just haven't gotten the chance to digest them yet. Also, Plutarch's Parallel Lives. They aren't "history" in the sense that we know it, but they do teach you that, yes, the lives of famous ancient figures do have lessons that can and should apply to you.

After I passed a course and read other books on Algorithms, what really clicked Algorithms for me was this little known book: A Common Sense Guide to Data Structures and Algorithms by Jay Wengrow. This is what made Algorithms and Data Structures make sense to me.

Later I read DPV and it was great, too.

My reference to it only a month ago [0].

:)

[0] https://news.ycombinator.com/item?id=35621692

It was a technique of studying in a book, not the content, In "Introduction to the Theory of Computation" Sipser precedes every proof with a "proof idea" basically examining meaning of each statement before the "pencil pushing". You wouldn't miss the forest while watching the trees like this.