To be clear, this is not a ranting post. I have never published a book, but recently I have been wondering why are authors writing maths books that is extremely long, say, 600-1600 pages, and the inclusion of exercices makes the question more complicated.

Indeed, if a maths book does not have any exercise, then we can somewhat suppose that it serves as a reference book and the book won't play a big role as a textbook. For example nobody complains the lack of exercise on EGA.

But in my opinion, if a maths book includes exercises, it automatically qualifies as a textbook. So I wonder, when the book is extremely long, what are the author expecting? Finishing a book of < 500 pages within one or two semesters can be feasible, but for a book, rather advanced and extraordinarily long, like Hatcher's Algebraic Topology, Bump's Automorphic Forms, Evan's PDE, Görtz-Wedhorn's Algebraic Geometry (1600 pages!) and many other books that I can't name, the reading of the book can already be extremely time-consuming.…