I'm currently the maintainer of a smallish-to-mediumish sized open source project. Lately I've received several requests/ideas to write a book about the project and get it published. It sounds like something that would be worthwhile, and I have an interest in doing it. But, I have a fear that the 'real world' of publishing may just be a lot more problems and pain than I'd want to deal with.
I'm looking for advice on perhaps getting started, and information that would be needed down the line.
My project is "pygame" (www.pygame.org) and it is a library for creating games in python. Both these topics are "in demand" enough lately that many books are being created on each of them. Several of my users think a book would be a great idea, and I'm quite interested in the idea.
The first thing that really concerns me would be finding an interested publisher. How can I estimate what sort of market there would be for a book like this? What level of interest must there be before a publisher will decide to launch a book like this?
I've been told the reasonable way to get started is create an outline/table-of-contents for the proposed book and then see if that gets anyone interested. I've certainly read and gone through many technical computer books, and while it's easy to criticize over them, I can imagine it's an entirely different perspective when actually writing the book. I think I would know how to make the book work well, but it's hard to know if those plans would really come together. (I don't believe people set out to write a bad book)
Currently I have put together a good selection of tutorials, examples, and reference documentation for my library. This is one part of the project I do enjoy doing. I believe writing a book would allow me to "fill in the holes" and really put together a great guide on how to get started.
One thing my project really attracts is a lot of "python first timers", who decide to try out the python programming language on a fun little project. I believe a book would certainly need to be focused towards these types of users. On the other hand, there are plenty of good "learning python" books available on the shelves. I'd want to do as little of the language introduction as I felt was necessary. Perhaps publishers would be more inclined to focus the book towards more intermediate/advance users?
Anyways, in the end I'm very interested in hearing the stories of anyone who's been down this road before. Are the publishers decent people to work with, or is it "cover-your-ass" from start to finish? Would I be better off starting with the "big boys" or instead finding some smaller, more independent publisher?
I think you should start by reading about Philip Greenspun's adventures with publishers and also MathWorld's salutary lesson.
I'm in a similar position to you and am toying with the idea of writing something cohesive instead of just answering questions piecemeal on mailing lists. (I have a theory that doing so might even get me some thanks instead of the usual "your tools suck" flames!)
As a Greenspun disciple :-), my theory is to just put it up on the web and worry about the dead trees people later. Your readers will get to read your writing, you'll get the recognition, and there's no money in the technical book author game anyway.
I'm thinking write it in TeX using constructs that are easily HTMLified and write an HTML converter too. Then it'll look good on the web, and you already have beautiful printed copy.
But it's the writing that's the hard part. I don't know about you, but for me writing is like squeezing blood from stones.
The words roll smoothly and soundlessly enough across the page; it's getting them to flow down the arm, it's squeezing them out through the fingers, that is so difficult.The Blind Assassin, Margaret Atwood
Most publishers have author's guidelines that tell you how to approach them, and importantly, how they will accept a manuscript. They will send it to you if you write a letter to the publisher's office. Some of them are online, for example O'Reilly's Writing for O'Reilly.
I believe O'Reilly will accept submissions in DocBook, since they had a big part in originating it.
Magazines also have author's guidelines, for example see the Linux Journal's Author's Guide.
Consider, though, marking the whole thing up in DocBook, licensed under the GNU Free Documentation License, and posting the XML, HTML and PDF on a web page. That's what I'm doing with some of my writing. Some publishers, notably O'Reilly, will still publish GFDL books. See Why Free Software Needs Free Documentation.
As an out-of-work no-hoper with a code rage on, some web stuff and some journalism behind me I was smugly thinking along similar lines. The "no cash in it" thing is a bit of a twister though. Shame on my lack of artistic devotion to the true cause of starvation.
Of course if you get yourself a killer app and make it complicated enough or find yourself a popular subject and take it from a good angle then perhaps both food and karma can be yours?
New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.
Keep up with the latest Advogato features by reading the Advogato status blog.
If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!
del.icio.us
Digg
Google bookmark
reddit
Simpy
StumbleUpon
Furl
Newsvine
Technorati
Tailrank