Ask HN: Why are people still publishing bad math textbooks?

For those who only read through one book per topics, you are not the audience. Because at least you don't seem to be able to judge whether a book is good or bad. I dump any book as soon as I notice that it is bad, usually within 30 pages, and find another one.

I personally used 4 books to learn Linear Algebra last year, and ever since then I hated math textbook so much that I think I will never read any of them again, despite the fact that my HDD is full of math books. Because those authors usually leave blanks on somewhere maybe it is important, and I am so nervous to fill in any gap that may exist.

Some books are just objectively worst written. Like famous Strange's linear algebra textbook. That seems easy at the first glance, and then the learning curve is just skyrocketed to nowhere a beginner can make sense. I mean, how would someone teaching LU decomposition in the INTRODUTION part of the book???

And many recommends 3B1B's linear algebra series, but I don't think that is a great material. He is a great animator but definitely not even a good teacher. Those who experiencing the same frustration watching his videos may understand what I mean.

The problem of major math textbook is that notorious "leave it as an exercise" or "you should prove it yourself". I am here to learn, not "self-taught" self-taught. If I knew how to do that, then I don't need you. I need you (the book) because I want to know. But in order to learn a thing, first I need to learn how to prove. Then I need to learn how to prove a proof. Then again learn how to prove a proof of a proof... You get the point, it is a dead loop that is so obvious to any programmer.

Yes, I even read logic and set theory book and it told me to prove everything myself. But somehow I don't know if it is a typo or what, the LOGIC BOOK is showing wrong thing to me. So bye-bye. Typo is fatal to a math textbook.

There are few books that I don't hate thus far.

One is the Mathematics of Quantum Computing from Springer written by a Dutch guy. He proved that yes we can have a book that is so comprehensive and "author-proof". There is none unproved theorem inside the 800 pages book. All steps are printed there explicitly.

And Stroud Engineering Mathematics is exceptionally best written for the narrative and educational approach.

And maybe the Openstax Calculus. Author provided all important proofs and for those omitted, I never feel any loss.

Especially in mathematics, books are written in Latex, mostly words. Only words are in the book. I wish they can draw more venn diagram and flow chart or something like that, to augment the intuition.

I think it is non sense to set up barricade against learners. Every non harmful knowledge should be accessible to anyone and this will aid the society and the world also reduce climate effect (no one ever need to spend 6 hours googling to find out the solution).

So IMO making math harder than what it is and what it should be is basically pointless and it is causing damage to the humanity. Those professors thinking the opposite should stop writting anything.