Yes, I think if you are writing a technical book, it'd be better to focus on a more stable technology. That or plan for a revision or two after the initial release.
And I can tell you, after I finished my book, I didn't want to think about the topic for a month or so. Can't imagine committing to a revision or two.
Doing revision of a book is OK. but the issue is, when we start writing a book, the version may be say 0.2, and when we release the book, the version may be 0.4. So the main thing is, the book got outdated in the release itself. So just think about our situation. Do we have motivation to write another book? I had this situation and I have changed the content of the book at the end of final edit. This is painful.
That's a good point. That would be deflating to have a book where the code examples wouldn't work with the latest stable release. I guess this is where a self published ebook would be a better fit (plus automation to make sure that your code samples were up to date with the latest runtime).
I guess it depends on how much the technology evolves too. I think of my own experience in writing a book on cordova. The amount of work to revise it for new editions would be, I estimate, about 50% of the effort of a new book, just because there was significant overlap between the versions.
And I can tell you, after I finished my book, I didn't want to think about the topic for a month or so. Can't imagine committing to a revision or two.