Running Linux, Version 5 Behind the Scenes

Last year, I had the opportunity to participate in Running Linux 5th Edition. Mae West once said, "Flattery will get you everywhere". So, I felt flattered when asked to re-write the first chapter and add a major section in Chapter 28. For aspiring writers and interested parties, you might benefit from my experience with what we called RL5.

My editor at O'Reilly, Andy Oram called early one morning to ask if I had an interest in doing some writing for hire. I don't know any free lance writers who would turn down an assignment, so I accepted. The next morning, the guy from brown delivered an overnight package with a contract.

I have a subscription to O'Reilly's on-line library known as Safari but I wanted the latest copy of "Running Linux". So, Andy had a copy shipped to me. He also warned that the new version would double the size of the last.

A review of Running Linux, Version 4 came through the submission queue here at LXer with much praise. That put a curious spin on RL5 for me. I wondered about publishing a newer version of one of the most popular Linux books on the market while it had brisk sales.

That didn't matter much to me when the assignment came through the email. Andy had supplied what we might call a statement of work and the internal copy of Heterogeneous Networking.

When you see the finished copy of a book like Running Linux, Version 5, it bears no resemblance to the document we writers see. I received a XML document written in a text editor using OpenDoc. We do all editing and tagging manually. I consider working in that way a real challenge.

We not only write the text you see in the final copy of a book, we also have to tag every sentence and phrase. Some of the tags look like these:

< simplesect >
< firstterm >
< sect1 role="" id="runux5-CHP-31-SECT-1" label="31.1" >
< figure id="runux5-CHP-31-FIG-1" label="31-1" >
< graphic fileref="figs/fig31_01.tif" / >< /figure >

We submit our writing to Andy with all the terms tagged and he looks at it like the code in the Matrix. Somehow, he only sees the text. He has an empowering way of explaining his edits so as not to discourage the writers. Some might call it ripping your heart out but Andy coaches well.

Andy and I have some conflicts regarding writing style. Andy likes an informal style with some passive voice, apostrophes, the use of gerunds and sentences that end with prepositions. I go with my training in English Lit and composition. I defer to Andy since he's had long strings of successful books beginning in 1992.

I don't remember any of my professors publishing or winning awards in composition. Still, I think Andy has a weird way of writing. But, I would never tell him that since he has such a successful track record.

I made the above observations so you might understand why my chapters wind up with six or more rewrites. I don't mind having to rewrite every chapter six or more times. But once, I caught Andy editing his own edits and I have yet to let him live it down. In fact, I use it occasionally to salvage something I want to keep.

In addition to the XML format , O'Reilly provides templates for Openoffice.org writer. I prefer using OOo since I can see the chapter as it would appear in book form. Then, we tag everything in the template using the ORA styles and no one from the outside would recognize the drafts.

If the tags fail to work, our materials come back to us for proper tagging. The tools group wants everything perfect so they can make the conversion from whatever format we use to the industry standard: Framemaker. Andy does a lot of the tagging but likes his authors to fix the tags so we'll learn.

Before Andy submits his materials to production, he puts us through a painful process called technical review. In this phase, a team of experts in various disciplines review our work and make extensive observations. We then have to consider each observation and fix them. That starts a cycle of more rewriting.

When everything seems perfect, Andy bundles up everything and goes through some unrevealed process which returns something called "proof galleys". I hate galleys. By this time, I can't see anything. Something happens to all writers neurologically and we just do not see any errors or points to change.

My secret involves busting up the PDF in a Scansoft product. It takes the PDF and converts it to a Word format. I then run spellcheck to find some of the errors. That only works for the small stuff.

To perform a good job on the galleys, one should use a good ruler. With the appropriate ruler, one with a metal edge, a writer can review the galleys in approximately ten to twelves times it takes to write the material originally.

What does all this do for the book? While it seems somewhat chaotic to me, Andy has some method to his madness. (I know it's a cliche). The book comes out beautifully.

I recall receiving my author's copies of Running Linux, Version 5. I began reading it and now understand why people love it. Heck, I love it.

But, I'll tell you something you should think about. (Using Andy speak.) Right now, Running Linux, Version 5 kills the competition. We put out the best book in the Linux space when we finished that book. Don't blame it on Rio - blame it on the guys in Cambridge.

O'Reilly published one of my parts of the book called Powerful Remote X Displays with FreeNX. I invite you to take a look. I did not think it would turn out as nice as it did.

Finally, I hope this helped you understand the process of creating a great book. Maybe you will understand why Andy Oram receives kudos in the acknowledgement section of every book he has edited. I feel especially blessed to know he considers me a friend. He's my hero.