I can sort of understand wanting paper reference books, since it can be nicer to just flip around instead of googling, but it seems silly to complain about recency, since that's the _obvious_ tradeoff in getting a paper book instead of just googling for the wc3schools page on the <section> tag.
Tutorial books are a tougher one, because it's really hard to nail the audience. Is it really for beginner beginners who don't even know syntax, or just for programmers who are new to the language? Dummies books are for one group, the pickaxe ruby book is for another. I do feel like Oreilly has been nailing this recently, releasing a number of small but info-heavy tutorial books on things like stats and machine learning, while also pushing out their bread & butter 800 page language/MS Office books, and the Head First series for people new to the programming world.
With new versions of CRLS and Knuth in the past year or two, the fundamentals world is doing ok, although since they're primarily textbooks it may be a while before we see them in ebook format =(.
There is a big difference between the good and the bad in this realm. A good book is worth keeping even after you've passed the exam.
I wonder if something like the pull requests of GitHub could be meshed with a literate programming tool (for easy extraction of source code from document) to produce the ultimate, sustainable tutorial infrastructure
E-book has far less advantage over paper when the material is being consumed sequentially.
I was the main author on a For Dummies book and the whole point is to provide step by step actions to guide a complete newbie through the procedure. It's intentionally done that way because that's what the audience wants. They don't want to read it cover to cover, they want to pick up the book, open it to the section called "Installing a wireless network card for Windows Vista" and follow the instructions. When they want to read sections, they want to read something that speaks to them, not something that assumes they've read it all up to that point.
Technical books may have failed a certain segment of the population (I agree with many of his points if I look at it from my own perspective), but I know a lot of technical authors who make a fine living writing books, which means that people are buying them.
Linking content in the books to outside real-time resources.
Alternatively, what would you think of a subscription based purchase model for a book where each year you pay $3-10/year for a book which is continuously being updated?
That being said, technical references have two roles, one to provide a reference 'now' and one to provide a reference 'then'. I could imagine that there is a market for an 'ebook' type thing which is a combination 'git repo / document' where you could read it like a reference book on your ebook reader, but you could slide a 'timeline' type slider which would assemble it from the document as it appeared at that time. Would be a fairly complex document, but really really useful.
