Thanks for the Wakeup Call! Jeff Atwood Explains Why I Shouldn't Write a (Traditional) Book

Jeff Atwood's Coding Horror is one of my favorite blogs.

He recently wrote about his new ASP.net book, and instead of telling everyone to go buy a copy, he spent his words explaining why one should not write a technical book.

To summarize, he pointed out

1. The format and process was painful to work with
2. "Writing a book doesn't pay"
3. One shouldn't associate any extraordinary credibility with publication. Jeff's a little more blunt: "Anyone can write a book. ... The bar to publishing a book is nonexistent..."
4. "Very few books succeed ... In the physical world of published atoms, blockbusters rule" or, in case you forgot the pre-web-2.0 world, the long tail isn't supported by the economics.

These arguments hit home with me because I had been preparing a book "proposal proposal," and chatting with some publishers (I say "proposal proposal" because a proper book proposal has a certain form and content which I have not fully developed; what I have is a more informal proposal; i.e., a proposal for a full proposal).

At a certain level, I already knew I am living in the wrong decade for the dead-tree book. But I also imagined that, since my topic is a little bit more process- and project- oriented, rather than "how to write an app with the new foobar 3.0 API," my book might have more long-term relevance or staying power.

But Jeff's post hit me like the knocking at the door in Macbeth. He's right. At the end of the day, the publisher and bookstores are not going to do anything impactful to promote my book; it will be hard to find on a shelf with 1500 tech books that change every 3 months, and I wouldn't expect it to sell remarkably well.

My main goal in writing is to share some knowledge and experience with folks who are interested and who might benefit. And writing and publishing online, I have a lot of confidence that my audience will find my content.

This belief comes from looking at the web analytics for my blog: through the magic of Google, I get solid traffic for meaningful keywords. For example, my post on the .net implementation of J2ME hidden (?) in Y!Go, but which you can use to port your own MIDlets to Windows Mobile, is result #8 (today) if you Google "j2me .net implementation" ... and it drives a bunch of traffic.

That's good enough for me.

Now I'm realizing that there are no advantages to the legacy tech book publishing process for me (I'm sure that for established, famous authors who write for a living, it's a different story). But there are a lot of advantages to avoiding that publishing process. Beyond the simple advantages of being accessible via Google, etc., I will surely save an enormous amount of time interacting with the machinery of the publishing industry.

Time saved is time I can spend writing and refining my core content. And if it's not perfect, that's ok, because there's no offset printing and physical distribution. I can fix typos or code errors instantly. I can revise and re-post if I realize I've said it wrong. I can offer old versions if anyone cares to see them.

Meantime, if anyone wants a printed and bound paper copy, there's always lulu.

Thanks, Jeff!