Ask HN: How do you write a book? | Hacker News Reader

Ask HN: How do you write a book?

33 pointszeynel26 comments

What is the best way to organize a complex or chaotic topic into an organized whole? I'd be grateful for practical solutions people actually used to complete a book. Thanks.

View on HN

The advice I've gotten: Write a book proposal. In it, lay out chapters with one or two paragraph summaries. The process forces you to think a lot about organization. It also provides a blueprint/outline for you going forward. You'll see how the ideas fit together to recognize missing gaps in your thesis. Keep in mind that it's an evolving document and as you write more you'll be better able to incorporate new material and scrap old forks. Starting with the summary provides clear signposts along the way.

Good luck!

Edit: Of course, I forgot the reason why many folks also write a proposal: To shop it around, get an agent, and an advance. Problem is, that usually requires some connections in the industry, special expertise, compelling story, and/or all of the above. You don't need to shop it around to get the benefits, though it's nice to have if you happen to run into someone who could help.

I've just finished writing a book that will be published by O'Reilly next year. I'd echo your comments:

1. Write an outline. This is boring, but vital. If you don't do this you won't realize that extent of the task you are taking on. By writing an outline you are convincing yourself and a publisher that you can actually write a book.

2. Do some other writing first. For example, I wrote extensively for different magazines. This is both good practice and gives you sample writing to show to a publisher.

3. "Phone a friend" Do your best to find someone you know who can introduce you either to a publisher or to an agent. This also applies to getting VC funding, it's best to find some way in via someone your know (or someone someone you know knows) than just trying to cold call.

4. If you do get a contract track carefully your progress (I used a spreadsheet) so that you stay on target.

5. Do not get hung up on the right tools to use. I've written plenty of stuff in a text buffer in emacs, or in Word, or using a specialist XML editor. Your publisher may actually ask you to use a specific tool, and the reality is that the tool is the least important part of the process. Figuring out what to write, and then writing the sentences is hard. (I'm not joking about the last part)

6. Read your work aloud. After I'd written a section I would read aloud what I'd written to see how it flowed. And I'd modify, read aloud, modify, read aloud until it seemed smooth.

What is the subject of your book?

How did you get O'Reilly to accept it?

I've been thinking about some cool books I would like to write for O'Reilly. I live 10 miles away from their headquarters, which might help me.

Are there any other north bay hackers out there?

1. It's a travel book for geeks

2. I followed scrupulously O'Reilly's guidelines on submitting an proposal. I went through their normal channel for proposals and then I asked a friend who's an O'Reilly author to give them a prod for me.

3. I worked with the editor I'd been assigned to convince them that I could pull it off.

I'm going through this right now for nanowrimo. ;)

At first I used word (very lame, I know), but a friend recommended Scrivener (http://www.literatureandlatte.com/scrivener.html) to me a while ago, and I've been using that instead, but it's an OS X only tool. I love it. I'm not a writer, but I really needed some tools to help me finish nanowrimo, and scrivener is it.

I've occasionally used mindmapping tools for occasional papers I've written, if it's extremely chaotic you may want to look into that, although I don't know how much help it might be.

Writing one blog post per day on Overcoming Bias, having to do that every day and publish immediately and put it behind me, and then getting feedback, has sped up my writing by over an order of magnitude.

Gerald Weinberg has written a wonderful book "Weinberg on Writing" http://www.amazon.com/Weinberg-Writing-Fieldstone-Gerald-M/d... where he outlines his "Field Stone" method. It's an approach that likens writing a book to constructing what the Irish call "dry stone fence." You write capsules and modules that you have energy around and then gradually re-work them into a narrative once you have a good understanding of the topic and how you want to proceed. He also ha a blog at http://weinbergonwriting.blogspot.com/ devoted to his thoughts on writing. If you are unfamiliar with him, he is a bestselling technical author (and now fiction) who has written more than 40 books.

Some advice that hasn't been mentioned yet: make sure your publisher supports you. Writing a book is difficult and you want to spend your time on the content -- not making sure that there are no typos, teaching the publishers how to run the code, doing the layout, maintaining different versions of the book, etc.

I recently wrote a book, and I got very little support from the publisher. It was a disaster, although I didn't realize that until after publication. I was the first person to write a Perl book for them, and they had nobody on staff that could even install Perl. The editors didn't do any significant editing, and in fact weren't even native speakers of English. They didn't even manage to publish the right version of some chapters. When they made the code from the book available, they did so as a word document containing code snippets from the book (instead of the runnable tarballs from each example and chapter that I provided for them).

Basically, your name goes on the book, so any fuck-up will be attributed to you. If something goes wrong, like things did with my book, you will get hate mail every day, criticizing you for things that you had no control over.

Oh yeah, and when the code you wrote the book about goes out of date, people will send you hate mail for that too. It is also apparently a hobby of people to nit-pick things like your variable names ("OMG you used $c instead of $ctx, YOU ARE SO DUMB!"). I'm not sure what motivates people to send me e-mail about things like this. Do they I think I care what they think? Do they want me to magically change every copy of the book? I don't get it.

In case you haven't noticed, the experience has changed my life. Before writing the book, I was really excited about open source projects and talking to people. I genuinely wanted to help people learn to program better. Now I am bitter and misanthropic because of the constant influx of hate mail. (I could go on and on about this, how people don't want to learn, how teaching is a waste of time, etc., etc... but I will save that for another day.)

I'm not trying to talk you out of writing a book, of course, but just make sure that your publisher cares about your book. If they don't, you may regret writing it.

I think what Jon's trying to say is "make sure your publisher doesn't do what Packt appear to have done to the first ever book on the Catalyst framework".

Not only did they successfully kill jrock's book but based on his contributions to the project and associated work they also killed his motivation.

So: good publisher, thick skin, pick two or don't start.

Indeed. But you don't need a thick skin, you need armor.

I wrote a novel last year. For fiction, it helps to just start writing. It's much easier to revise what you've got than it is to start writing anew, so when you have time, just write. Get your ideas out. Once you have a lot, then you can begin to revise: get rid of what isn't absolutely necessary, make sure all of your sentences gleam. Revision isn't always fun, but it's much easier to get in a revising mood than it is to get in a writing mood

If you're writing nonfiction, I'd guess that you'd start by researching whatever you're writing about. George Carlin once said that as an older man, he could make logical connections in topics he could never talk about before, because his knowledge was more comprehensive. Same with nonfiction: the more you know, the more ambitious your writing can be. (Same for fiction, of course, but with nonfiction you're more often writing about a specific topic.)

In both cases, outlines help a lot, but it's possible to work without them. It depends on what you want to write, and how you go about writing it.

"Just start writing" is a fiction thing.

The nonfiction equivalent is "just start writing index cards."

I always like to start with a mindmap (http://en.wikipedia.org/wiki/Mind_map) of the key points. Once I have those, I flesh out the individual nodes underneath. I then organize into Chapters/sections and start writing.

I've read where some people create a PowerPoint presentation and work backwards from that as well. The key thing for me is that I don't want to get tied down to Chapter 1, Chapter 2, etc too early - I want the flexibility to move things around without hindering my thinking process. That's the sort of thing that is perfect for mind mapping tools like FreeMind (http://freemind.sourceforge.net/wiki/index.php/Main_Page).

Just start writing, using whatever works best for you. I've been using TextEdit on the mac a lot lately for my writing, but anything's good.

If you're writing for a publisher, then they're going to give you a template that's probably for Word, and you don't have to worry about complicated document struturing stuff -- they'll help with the referencing, etc.

If you're doing it all yourself, then a document description system such as LaTeX might be the way to go. If I was writing a technical book that i knew had little chance of being published, I'd probably use that.

"Just start writing" is pretty terrible advice for writing a non-fiction book. The greatest value of such a book is in the synthesis of the subject at hand... without high level organisation, that synthesis simply won't be. Chaotic, messy non-fiction is terrible.

Fiction writing, on the other hand, can go totally chaotic... but I imagine the OP is talking about non-fiction.

Actually it is ok advice taken as the first step. Write to see what you think. Then rewrite, rewrite, rewrite.

Perhaps this is one of the situations where "build one to throw away" is good advice. Once you've spent four weeks writing a sketchy book-length ramble on your subject of choice, you'll have a much better idea of the size of the task ahead of you, and plenty of ideas on how a much better book on the same subject would be arranged and written.

Then, don't start rewriting your manuscript. Burn it and start all over again.

I've never written a book, but I plan to eventually, and this is the approach I plan to take.

I actually wrote some 150 pages of a first (technical) book before sending out proposals to 10-15 publishers to which I received a resounding ... nothing. I think maybe one rejection.

After a year of feeling a little bummed about it, I tried again, this time pre-clearing topics with the publishers and got a proposal accepted.

But the advice that stuck with me is to just write. Sure, plan out what you're going to write ( topics + sub-topics for non-fiction, storyboard for fiction), but otherwise, just let the words flow. You're going to butcher the crap out of it later on when you start editing it anyway, so don't _start_ by worrying about it.

I still have those 5 chapters of that first "book" I never wrote. I like to go back and look at them every once in a while -- they're not great :-P.

An "old" method used to be to put individual items on notecards. These could be rearranged repeatedly, with additional cards inserted as ideas were either fleshed out or fit into a larger context.

It's writing, but not (yet) writing a linear, not to mention final, draft.

I understand that these days some people work in a similar fashion, but with Post It notes. The advantage being that you can spread them out over a table or wall or similar (walls are then a convenient source or otherwise unoccupied surface area), and they won't be as easily inadvertently jostled out of order.

If you are worried about losing a card organization scheme, when they are laid out in such fashion you can take picture(s) in order to preserve a record. This can also serve as a backup mechanism against losing the cards.

It also works for handwritten notes. You do not need to spend a lot of time on the pictures if that time is not productive. But the pictures are so easy to take -- one can just snap them at the end of each day or workday. If your 4 year old randomizes your stack, or you forget your notebook while juggling 3 other things, all is not lost. If that never happens, all you've lost is a couple of minutes' clicking and some non-volatile memory capacity.

Behind all these comments, is the fact that for me, writing long hand is still sometimes most productive. If the volume of output is less, the concentration tends to be higher. And it can be accomplished most anywhere, at a moment's notice.

There's also the kinesthetic aspect of long hand, to which I respond strongly. Knowing one's learning and interaction style may be helpful. For me, a physical relationship/interaction drives heightened awareness. Others may be more visual, or acoustic. I wonder whether personal recorders / voice notetaking devices work particularly well for the latter.

I pretty much just sat down and started writing. I skipped the planning and just let it flow. That meant a good amount of cutting and pasting things around later to get everything in the proper order, but the free form inspired me to write more, better, and faster. Your mileage may vary.

Do you have a contract and an agent yet? I learned the hard way the value of a good agent.

Thanks everyone for great advice! This was very helpful

Understand it.

Notepad

The best way to write a book is to work on it every day.

Reply on news.ycombinator.com