I can now report that I'm within spitting distance of the end of the first draft of Flight of the Vajra. Don't pop the bubbly yet; there's a long way to go.
I've never written anything this long before — 350,000 words or so — and so I had to change my work habits in ways that were humbling and frustrating.
One major change I made was to allow myself to add in-line notes to myself during the writing process, via Word's annotation function. If I was in the middle of a sentence and told myself, "I should make sure the foundation for this is properly established", or "I suspect we can put this somewhere else", I'd note that right in the document and keep going. The hard part is now figuring out how to go through all those annotations and apply them. Do I just do that front to back as I work on the body of the text itself? Or would it be best to just clean up what I see first, and then apply any annotated changes, in effect making two separate passes?
To complicate things further, there's my sidecar wiki, in which I tracked details about the story (characters, plotting, locations, etc.) Several different sections in the wiki would also need to be trawled for possible changes in the next drafts:
The "Goodie File". This is something I picked up from reading The Illusion of Life. The animation teams at Disney, while working on a particular project, would keep a pin-up board in the corridor outside the animator's offices which they called the "goodie board". On this they would pin ideas — sketches, concepts, scribbles — which seemed good, but really didn't fit into the story as it was. If one animator was tossing out an idea, another one could walk along and pick it up, and perhaps it would serve as inspiration to solve another problem he was having. In the same spirit, I reserved a space for jotting down things that seemed interesting — a scene idea, a gizmo idea, etc. — but which didn't have any particular context or connection to anything. Later on, if I was stuck, I could poke back through the Goodie File and see if its contents sparked a connection. They often did.
The "questions and answers" section. This was a portion of the wiki where I'd pose to myself unasked (or unanswered) questions about the things going on in the story, as a way to challenge their assumptions or break up stalemates. Many of those only got integrated into the story in the sense that I would continue writing under the premises provided by the answers I'd picked at a certain point, but I'd still have to go back and make sure everything earlier also matched.
The "future drafts" section. This is more or less what it sounds like: an area to dump things I know need addressing the next time around. They're not in any particular order (it would be difficult for them to be), so I'd probably need to trawl through that file on its own as well.
The "plot flags" section. Relatively short — anything I started and told myself I would need to finish would get annotated here.
Something I noted to myself for the future was how a good deal of the wiki's sprawl could be eliminated. I'd created way too many subdivisions for things, out of a sense that the more precise I had a place for something, the better. It turned out that I only needed a few places, but I needed a precise workflow for each particular piece. Handling ended up being more crucial than just organization.
Example. The above files could easily be condensed into two things: the Goodie File and the Future Drafts section. Q&A could hold one question at a time, and when that question was answered its answer would be turned into an annotation for Future Drafts, or dealt with immediately.
Another big lesson I learned from working on the story this way was that while a sidecar wiki was indispensible, it also created its own traps. Chief among them was treating the wiki like a dumping ground, instead of an organizational tool. "Hey," you tell yourself, "it's a computer! And computers a great at freeform data organization, aren't they?" Yes, but only to the extent that humans program them properly.
So my ultimate approach will probably go something like this:
It's the last part, actually — getting distance from your own work — that's the toughest. But that's why we have friends we can bribe with offers of dinner on the town for reading our work, isn't it?
(And yes, Gabe, I still owe you a read. You Are Not Forgotten.)