Online Book Publishing Workflow?
February 23, 2012 3:56 PM   Subscribe

Online Book publishing: ebook, kindle, epub, ibook - can this be done without a million different programs and converters?

I have an existing book I've converted to one long html document (HTML+CSS). I've laid it out in html, and placed each chapter as section elements.

What's the best practice for turning this into a book suitable for publishing on the various bookstores. The idea is to publish this everywhere if possible: Amazon Kindle, Apple iBooks, Google eBookstore, ad-hoc sales. PoI'd like to have some degree of control of the output.

There are a JILLION sort-of-okay, sort-of-out-of-date tutorials on doing this in pieces. I also see Calibre looks like it can do some of this in conversions. But I really don't have a great sense of best practices here in terms of how to use links, how to use headers, what different readers do in terms of providing indexes or metadata (like, can I add meta tags that tell the title, word count, how to contact publisher, etc?), and how about links, I'm doing one large document, and am doing the TOC as links to id based anchors. Is that fine? Would I be better off breaking it up?

So I guess I'm asking for guidance on a workflow for making ebooks for the different marketplaces using the same "base" file to create the different versions.
posted by artlung to Technology (13 answers total) 20 users marked this as a favorite
 
If you've got a book in HTML, I think that Sigil is what you are going to need to turn that sucker into a beautiful ePub.

After that, you can use Calibre to turn the ePub into a MOBI, which is what Amazon uses.
Barnes and Noble and Apple both use ePub as their preferred format.

Sigil is here: http://code.google.com/p/sigil/
posted by THAT William Mize at 4:47 PM on February 23, 2012 [2 favorites]


Seconding Sigil. And you really only need to render it in Kindle format (.azw, which is a proprietary variant of MOBI) and in ePub for the US.
posted by Sidhedevil at 5:04 PM on February 23, 2012


Oh, I spaced out entirely--I think Amazon requires that all books for sale via them be in the new Kindle format. Using their tools is reasonably easy once you have a clean ePub.
posted by Sidhedevil at 5:08 PM on February 23, 2012


Epubzengarden.com has some lovely Creative Commons EPUB designs available.
posted by nicebookrack at 7:19 PM on February 23, 2012


You'll have to reformat it in Word, but if you submit to Smashwords according to their specifications, they can get it into ALL the other stores on your behalf, plus their own store.
posted by JoannaC at 8:15 PM on February 23, 2012


Sigil is nice, but it doesn't seem to address my question. What it seems to do is allow import of an existing html file, but it then converts it and the associated styles, modifying my code in the process. It then locks me into using Sigil.

I don't need eBook styles. I don't want to format to the limitations of Microsoft Word, so Smashwords is out.
posted by artlung at 8:20 PM on February 23, 2012


Christian Cantrell has a really good and regularly updated how-to publish on the major etext platforms (Amazon, iTunes Barnes & Noble) Linky. He addresses exactly this.
posted by Ookseer at 9:36 PM on February 23, 2012 [2 favorites]


can this be done without a million different programs and converters?

Probably.

Obligatory disclaimer first: I work for an epublisher, and while I'm not directly involved in our production process, I'm aware of how it works and have selected some of the programs we use in it.

Basically, though, what you need is the cleanest epub file you can get--the places you listed all either require them (Apple, Google) or accept them (Amazon CreateSpace, which is, I assume, what you'd be using).

You can't use Sigil or Calibre to create the files--neither of them produce epubs that consistently validate, and (in my experience) neither iBookstore nor Google will accept files that don't validate. This also probably rules out using InDesign, unless you're willing to manually edit every file you create. (It's only a single line, if I recall, but it's still a stupid thing, and you don't want to buy ID for this.)

The best option that I've found is sort of an odd one--a company called Epingsoft sells ePub Maker. It's this unbelievably simple to use program that costs $50 and converts from HTML or Word doc to a fully validating epub. (I've got no ties to the company, I'm just very pleased with the product.) While I've not tested it with any especially complex layouts, it's done very nicely with everything I've thrown at it, which included a book with an extensive family tree, some with graphics, and some with tables. These have all been rendered remarkably faithfully, though it's worth noting that with one or two (simple) exceptions, all the files I've converted have been doc files. Links are kept intact and functional, in my experience.

You might have to alter the document's structure (I'm not sure if it accepts the section code, or if you need each chapter as a separate file), but it'd be mostly breaking it into pieces, I think, not totally reformatting. Metatags are supported--off the top of my head, I think that they support book name, author, publisher, description, ISBN, and language, maybe? There may be others that I've forgotten.

In terms of indexes and metadata--some readers and management programs seem to support metadata reasonably well and will allow you to sort by any of the data, and have it all easily viewable. Others only support basic fields. (Nook, Kindle, and Kobo can all be counted on to have reasonably good metadata support, though I believe that all three sometimes have issues with summaries/descriptions displaying for sideloaded files--that said, if you're going through the official stores, you're unlikely to run into this problem.)

Indexes are embedded in the epub itself--they're not created on the fly, if that's what you're asking. Every reader I've ever used (Kobo, Nook, Kindle, generic Android tablet, Xoom tablet, iPad, various other bargain-basement readers and tablets) displays the index without problem and allows readers to use it to jump from one point in the book to another, assuming that the index was properly created. (Many novels don't have proper indexes, which makes me batty.)

Hopefully this was helpful. It's rather late here, so feel free to poke me if I missed things or stopped making sense at some point.
posted by MeghanC at 10:15 PM on February 23, 2012 [4 favorites]


So, clean and validating ePub seems to be what I need to be targeting. So what does that mean. Wikipedia lists two different variants: 2.01 and 3.0, along with some other components that be part of it.

It looks like if I'm comfy with xml and xml tools (which I am), reading the ePub spec is in my future.
posted by artlung at 10:28 PM on February 23, 2012


3.0 is brand new--like, last couple months new, if I remember correctly. 2.01 is still the standard for most places, and I'd stick with that.

And yes, if you're comfortable with XML, definitely read the spec.

ePub validation is done through ePub Check. If you validate through that tool, you're likely good to go just about everywhere. Note that even things like "unable to decrypt" or encryption issues or whatever will return as errors when uploaded to vendors such as Amazon and iBookstore.
posted by MeghanC at 10:49 PM on February 23, 2012


If this is a one-off, Sigil or ePub Maker might be fine, but if you run into trouble, you should know how to edit the code. The reference you want is Liz Castro's EPUB Straight to the Point. It's slightly out of date re: the new Kindle tools, but there are updates on her website. Joshua Tallent's Kindle Formatting is also pretty canonical. Another resource if you get stuck: #eprdctn on Twitter.
posted by libraryhead at 1:45 AM on February 24, 2012 [1 favorite]


Just reread the question, and I noticed the part about the whole book in one file. Technically, this should be fine as long as it's not over the size limit (300 mb, if I recall), but Kindle won't honor page breaks in the CSS, so if you want your chapters to start a new page, the most foolproof way is to have each chapter be its own HTML file. Also there's an iBooks bug with intra-file linking that you'll want to avoid; inter-file links work as expected. (sorry for typos, doing this on the iPad)
posted by libraryhead at 2:07 AM on February 24, 2012 [1 favorite]


libraryhead, thanks much, those are killer links. And Ookseer, the link you provided is epic in what it provides, though I wish he actually shared his scripts, though I can guess what they do in terms of creating the associated xml files and bundling them as a structured zip named as an epub.

It looks, to my eye, that EPUB is as much a standard as HTML is. Which is to say, there are common bits, but there's a lot of bugginess as well. This is definitely an education in the realities of EPUB creation.

It would be helpful to know if there tools to help test an EPUB on different devices as well. Simulators or services, etc.
posted by artlung at 10:09 AM on February 24, 2012


« Older Lunges make me lose my lunch   |   Will my mental health really improve if I start... Newer »
This thread is closed to new comments.