EPUB support

I’m throwing my virtual penny in the wishing well and asking for the next version of Scrivener to support an “Export to EPUB” feature. With EPUB gaining traction via Adobe Digital Editions and now the Sony Reader, it would make Scrivener the only dedicated writing package that I know of to support e-book creation.

Here’s a link regarding some of the technical details:

What exactly are you looking for Scrivener to do? Reading the description here, it can already do most of this—and the parts it cannot do would require a lot of feature additions. For example, generating the XML meta-data; Scrivener doesn’t internally store that kind of meta-data anywhere, so a whole support UI would have to be built for that. As for the XHTML book content itself, Scrivener exports XHTML in two different ways.

This sounds like a good application for MMD, really. There already is a meta-data area in Scrivener for that. It exports semantic XHTML instead of look-alike XHTML. And writing post-processing scripts to generate the XML meta-data file and zip up the package would be trivial to write.

Plus, using MMD means you can also easily generate to other, more established eBook formats as well. eReader, MobiBook, PDF, all of these have tools for easy conversion to specialised formats. You wouldn’t be stuck with just this one format that, from the looks of it, has yet to really be widely supported.

I haven’t delved into Multi Markdown just yet but thank you for the information. There is an open source project for OSX called the Daisy Pipeline that supports eBook publishing in multiple formats as well (daisymfc.sourceforge.net).

My hope was that Scrivener could be one-stop shopping for easily exporting to an eBook format that’s compatible with devices like the Sony Reader or the Kindle. From a workflow perspective, I use my Sony Reader in conjunction with a legal pad instead of my laptop for reading drafts. (Yikes! That sounds pretty self-serving. Sorry!)

I picked EPUB not because I’m particularly married to the format but because Adobe has a pretty good track record for instituting standards and eBooks.com is supporting it.

emediawire.com/releases/2008 … 156304.htm

The bit about “We’re keen to enable self-published authors to upload their books to boost the number of titles in our repository…” was the piece that really caught my attention.

Thank you for taking the time to answer my previous post. It was really heartwarming to see that you consider all wish list items.

A different Keith

Thanks for the link to Pipeline, that is indeed an interesting project. I’ve been looking for a way to go from Markdown to Kindle (which is just the Mobibook format, basically) without a bunch of run-around, so I’ll look in to this. I have tons of things archived in the MMD format, and it would be fantastic to have access to all of that on the Kindle. I don’t mind reading MMD “syntax” in a straight txt file form, but having table of contents and so forth—of course, would be nicer.

One thing you will find quite quickly, if you haven’t already, is that none of the standard Apple exporters (which Scrivener supports, though some are more advanced than the vanilla versions) have a clue about document structure; even the formats which do in fact support these things, such as ODF and XHTML. The Original Keith has put a lot of work into making these export engines a bit more savvy. They can do real footnotes, Word style comments, and so on, but anyone wanting to set up a chapter outline based on titles will still have to manually go through in a word processor and set styles and markings.

This type of structure is fairly important to the eBook format, as most reader devices take good advantage of footnote linking and table of contents layout. A book without these features is very difficult to navigate on a reading device that lacks rapid scrolling and in-document searching, such as with a computer.

Currently, the only reliable way to get that kind of structure in Scrivener is to use its MultiMarkdown facilities (some have used macros in word processors to convert standard Scrivener exported font styles into true styles). They are remarkably easy to learn—though of course any hand-tooling of the XML transforms and Perl scripts will jump straight into geek country (nonetheless, there are a number of us here are do just that); but for straight exports to XHTML and RTF, it is very simple to learn, and the resulting files are easy to read by a human even without conversion. No messy syntax.

Anyway; since Scrivener invokes Perl scripts as part of the export process, the possibilities are essentially limitless. Setting up pipeline projects and so forth shouldn’t be too difficult for anyone that knows how to put a little post-processing scripting together.