Publish or perish. We all know about perishing (or, if we don't, it's easy to learn) but publishing is another matter. Here is a splendid book that tells all.
And I do mean all. The book has 115 sections (and eight appendices, and a 30-page glossary, and a bibliography) some of whose titles are
etc., etc., etc.
The author knows whereof he speaks. He has written dozens of books and papers and has edited widely. His book is full of good and specific advice and should be read by anyone who is embarking on a career in academic mathematics. Even those of us whose careers are well underway (or almost over) can learn new and interesting things from it. Do you know what a Smyth-sewn book is?
He writes with clarity and wit. Even readers who have no need of instruction can read the book with pleasure, because good writing is a pleasure to read.
Another reason for having to book is to be able to look at it again twenty or thirty years from now. I'm sure that mathematical publishing then will be different in many unforeseeable ways, and the book will let us recall how things were done in the good old days.
Though the author tells us how to state a theorem, he does not go into the subject of how to prove one. Perhaps in the second edition?