Tinderbox And Proofreading

Web sites -- especially densely linked sites -- are often built up of lots of small pages. And with Tinderbox, even those small pages are often assembled from bunches of indivudal notes. This is terrific for writers and editors, but proof readers often prefer to work with a lot of straight text -- not dozens of tiny little passages. And they often have strong preferences for the medium they edit: some insist on double-spaced paper copies, others swear by Microsoft Word.

I can't read proof, so it's all arcane to me. How can I keep the proof reader happy?

It turns out to be easier than I'd thought. First, we add a new agent to the Tinderbox document. It gathers everything that's published in the site -- but none of the boilerplate headers and footers.

#inside(archives) & Prototype=WebPage

Now, we give that agent a simple HTML export template that includes each of the notes it finds in turn

^justChildren(proofReaderTemplate(

This goes through and adds each page, in turn, using a special template adapted for our proof reader:

<H2> ^title^ </H2> ^text^

This winds up creating a long, long HTML page that you'd never want to post but is perfect for importing into HTML. Apply a stylesheet so our proof reader gets the fonts, margins, linespacing, and everything else just the way the like it. Email to proof reader, or print it. You're done.

If we want, we can even add an information block in the proofReaderTemplate that adds metadata for the proof reader -- things you might want to know in house but won't publish. When was this passage written? Who revised it last? Has it been approved by the legal department? If the proof reader needs technical help, who is the right manager to contact?