ICE for lady novelists

[Updated 2006-10-06: fixed spelling of Westerfeld]

This week Justine Larbalestier tells you how to write a novel. That's not how I wrote my 2.4 novels – but then mine were not publishable.

(“Competently written but lacking the excitement our readers expect” said Mills & Boon. “You have not been successful” from the nice helpful people at the Vogel prize, who were all about encouraging young writers by giving them useful feedback.)

Obviously I'm in no position to comment on  Justine's advice, so I'd like to comment on the technology.

Justine says:

On that computer you need a word processing program. If you want to be compatible with the publishing industry it should be microsoft word. If you want a program that doesn’t make you froth with rage it should be anything other than microsoft word. (Sadly, I have gone with the rage-frothing option.) You’ll also need some kind of spreadsheet program which needn’t be compatible with anything else—it is for your eyes only.

Word aint so bad if you can get it set up right. The ICE project is all about that; we started with optimizing Word and OpenOffice.org Writer for courseware writing, and the ICE-RS project will look at optimizing it for research. I reckon I could use it for writing a novel without too much pain. But it was helpfully explained to me at age 28 that I  am “not successful” as a novelist so what would I know? Also I am unable to generate excitement.

Justine goes on to write about her husband Scott's method for managing a book in a spreadsheet.

At a glance I can see which pov was telling what chapter, what day it was, where they were, and who was getting the lion share of the novel. You can also have a content column that lets you know whether it’s a sitting-around-talking chapter (”) or a sitting-around-and-thinking (’) or an action-packed chapter (!) or somewhere in between (^) or one with sex (*).

This is a Good Idea.

What they do in the Westerfeld / Larbalestier household is keep metadata about the parts of their books in a spreadsheet and use that metadata to manage and plan the production process. I started wondering how this could work with ICE. Instead of putting the metatada in a spreadsheet why not attach the data to the documents themselves? Then you would go to your ICE web-view of your novel and see an annotated table of contents that showed which chapters are '“' and which are '!'.

(And why not add this information to the published version so people can find the '*'?)

I'm imagining that one would keep each chapter in a separate document and use ICE to put the chapters in order using the organize feature (making it trivially easy to change the order later). ICE watches what you're doing and turns everything into HTML, which you may or may not need. But ICE is good a presenting things in nice ways. How about little icons for the protagonists so you can see who's point of view is where in the novel?

While we're at it, how about:

1. Automatically generated word counts.

2. Dialogue to description ratio worked out by the computer.

3. Ratings from first-readers – give them a form so they can enter comments but also simple star-ratings.

I'm imagining a dashboard view of the novel, with green and red lights to show where the work is needed and suitable icons for Justine's categories of  talking, thinking, doing and rooting.

Cameron Loudon and I have talked about this for research workflows. We want to allow you to attach stuff to your documents about what journal or conference you are targeting, deadlines etc and then let a management system report to you about what you have due when. Or one could report on the progress of a group in aggregate. Imagine how that would look at Justine's place with two writers working on multiple fronts. Who's novel's going better? Who has the most words per chapter?

Or for a thesis, you could look at metrics like how many quotations and citations you have in each part, and ratings and comments from your supervisor.

ICEalso takes care of a couple of other things for a writer, the main one being backups and version control. Here on this blog for example I write on any one of a few computers, on a 'working copy' of all my stuff. When I hit the sync button it all gets sent to a virtual server I rent somewhere in the USA. That server keeps track of every version of everything I write.

(Justine – you do back stuff up, don't you?)

If I want to roll back to the state of my work on a particular day I can. This gives great freedom to completely refactor a whole chapter to see how it comes out and revert or roll back if it comes out wrong.

(There's a caveat here. ICE does not yet have features to do this, you have to resort to using the Subversion system directly, or get someone to set up a  web server application such as Trac that lets you browse your old versions via the web. There is a nice Windows interface called the Tortoise for Subversion. And the commandline works even better).

If you want to share a work in progress then your collaborators can hit 'sync' to get the latest snapshot. We're adding annotation and rating services over the next year that should allow us do what Justine is doing with her spreadsheet.

Using ICE you'd also get an HTML version of your book with all the required navigation. Some novelists like to put their stuff up on the web for anyone to read. ICE would make that really, really simple to do.