EPUB Me

There's consternation about the new .ibooks ebooks produced by Apple's iBooks Author and read by iBooks 2:

iBooks Author Requires Selling Original Books Through iBookstore, Says Apple's Controversial Contract – Jason Gilbert

The pros and cons of the iBooks 2.0 textbook format – Baldur Bjarnason

Will publishers feel trapped?

"Every publisher should have a publishing workflow." There is nothing novel about that. There is a base format (for the publisher) from which all publications are born. EPUB 3 is a candidate. Or something more general involving HTML5, XML, any combination of these with CSS and JS, etc.

The publisher might have their own EPUB with some additions (e.g. Apple's fixed-layout EPUB). The point is the base format is fixed, and from the base format all other formats flow. I'll just call the publisher's own base format EPUB Me, and the file extension .epme (though it might just be .epub following the appropriate specification).

There are various channels the publisher distributes through. These will likely require a conversion from EPUB Me to another format. Take Amazon's KF8 as an example. Thankfully, Amazon has provided a document and a tool for enabling the publisher to covert EPUB Me to KF8. The publisher first has to convert — how this is done is technology adopted or developed by the publisher — their own .epme to a .epub⁺ that KindleGen can accept. (That's where the document comes in: to define what type of HTML+CSS Amazon will accept either in .epub form or a directory of HTML+CSS+image files.) KindleGen will output .kf8.

Apple's format is not so simple: How do you go from .epme to .ibooks? Given the restrictions of iBooks Author, one way would be to use an EPUB to DOC converter (EPUBtoDOC^*, or E2D for short), import the DOC file into iBooks Author, and edit (hopefully without a lot of work) what is there to make it right for delivery in the Apple channel.

So there are two examples already of the publisher's workflow:

.epme → .epub⁺ → KindleGen → .kf8
.epme → .epub → E2D → .docx → iBooks Author → .ibooks

The key to this workflow is that the original source.epme belongs to the publisher alone and not to the channel on which it's distributed. output.ibooks may be "owned" by Apple, but source.epme is owned by the publisher.

iBooks Author seems pretty much like a black hole though. One might manage somehow to get EPUB in, but one can't get anything useful, like EPUB 3 (other than Apple's proprietary format), out.

I.e.: From authors/publishers point of view, they need to feed into the Apple .ibooks format if that turns out to be useful for them, but keep their original source (their .epme) separate, so they can feed that same source independently into different channels as well.

This workflow model can be extended to any number of distribution channels.



^* EPUBtoDOC: e.g. epubtopdfconverter.net/epub-to-doc.html. Alternatively, load the EPUB file into Sigil (EPUB editor) and cut-and-paste into iBooks Author. :(

E-publishing distribution formats for 2012

Here are the "container" formats for e-publishers to distribute in this year. There are a bunch of older ones which don't appear much anymore^* after 2011, The Year Of The E-Reader. (Some said 2010 was that Year, but it was really last year.)



EPUB (.epub)

Technically, it's all about EPUB 3.0 superseding EPUB 2.0.1. (I write "EPUB" instead of "ePub", etc. as that's what's on the standard.)

EPUB (.epub) fixed-layout

Apparently from Apple, it really just seems to be EPUB with pixel-specific settings in the "style" attributes (CSS) in the EPUB file. There are samples from eBook Architects. It's like PDF in that it doesn't "reflow". I guess whether one wants to call this a "different" format from EPUB is a matter of debate.

KF8 (.kf8)

Amazon's new format to "replace" its old MOBI 7. Could add to some confusion of publishing the "same" book in both EPUB (for NOOK, etc.) and KF8 (for Kindle).

KF8 released for publishers
Overview
For Word editors
For HTML+CSS editors

CDF (.cdf)

From Wolfram Research, this is the Computable Document Format (CDF). When CDF "players" are available for tablets, it could be huge in the math e-textbook domain.

How it compares to EPUB

HPub (.hpub)

Used by the Baker Framework to make HTML5-based books or magazines to be downloaded as iOS apps. Could be more powerful than EPUB 3 alone (e.g. allow more JavaScript-ing).

.hpub → Baker → .ipa (iOS "app")

As mentioned above, Amazon will "replace" MOBI with KF8 (sort of, and not yet for its E Ink Touch reader). PDF will stay around as long as people have printers (though fixed-layout EPUB replaces a lot of its non-reflow "features"), Flash will be edged out. HTML5 will rule.

That's about it. Any other formats worth mentioning I missed?


2012/01/21: On Thursday, Apple announced new format: a NotTechnicallyInEPUB (or "iBooks") format. More about when people can figure out what it means.

^* For conversion of EPUB files to older formats, or the conversion of HTML to EPUB, see Calibre (there is a list of formats there). Editing of EPUB files can be done with Sigil.

A comment on new eISBN recommendation

A new recommendation for assigning ISBNs to ebooks has been out a couple of weeks. I remarked earlier on why this didn't make sense to assign two different numbers to the EPUB ebook and MOBI ebook of the "same" ebook when the only operation was to take the EPUB one and use an automatic software converter (e.g. Calibre) to convert the EPUB ebook into the MOBI ebook. This is mainly done now simply because Amazon does not "read" EPUB, and the publisher has to produce a MOBI version for its Kindle customers. (In a way, this would be like asking a publisher to assign different ISBNs to a hardcover book depending on which bookseller chain store they sell that book in.) The EPUB ebook on the NOOK and the corresponding MOBI ebook on the Kindle "look" the same and have exactly the same HTML/CSS content. It doesn't make sense that they would have a different ISBN simply because one is reading that ebook on a Kindle vs. a NOOK.

The point is: If it's a basically automatic conversion that takes your EPUB ebook and produces that ebook in a different e-format, the ISBN stays the same. However, if one uses features of a particular format to really distinguish the ebook in that format (i.e., there is no simple, automatic way to convert it to another format), a different ISBN should be assigned.

(PDF is a curious exception. It is a fixed, not a reflow, format and is used for printing. There are converters of course that take an EPUB ebook and automatically creates a PDF file. Perhaps they would be useful to send one's EPUB ebook to a POD machine.)

Now there is a new format on the block: KF8 from Amazon. If one publishes an ebook that uses KF8-specific features that truly distinguishes it from the EPUB version, then it's a different ISBN. That makes sense.

CDF (Computable Document Format) may be an upcoming one for math ebooks.

Now some ebooks are being published as apps: iOS vs. Android. Here, the recommendation should be the same: If one starts with the same base version of the ebook (e.g. HTML5) and uses a tool that automatically produces the two different apps (iOS, Android) with the same ebook (HTML5) base, the ISBN would be the same. However, if the publisher uses iOS and/or Android dependent features to distinguish the two version, then it's two ISBNs.

Books in repose

I was sitting outside the Barnes & Noble cafe perimeter with my NOOK Simple Touch. I was using its Read in Store feature, which allows one to read any ebook available for purchase for up to one hour. You have to be physically present in the store with your NOOK connected to that store's WiFi, of course, to do this. At any time before the end of that hour, you can purchase the ebook or just leave it. But you could return the next day and continue reading the same ebook. In front of me were stacks of shelves of books — for convenience I refer to the traditional print editions of books as "books" and their electronic versions "ebooks" — and thinking of the future of publishing.

The first thing that struck was the irony of having to go to a location with books to browse ebooks. There were people in the cafe who were reading books they just took off the shelves that they were just "browsing" in the same way I was browsing ebooks on my Touch. (There were probably more people there with their notebook computers just using the store's free WiFi to browse the internet.) But I was there, as I had gone there before, just to browse (i.e. read) ebooks and I would not even touch a single book.

The second thing that crossed my mind were how alien the books sitting on the shelves in front of me appeared. It was like they should be in a museum, like ancient artifacts. People looked anachronistically odd walking up the shelves and pulling one out and looking inside. It was so much more convenient to read that book on my Touch. I could navigate through it so much better. In fact it was a better reading experience in every way.

I can understand, perhaps, some wanting art and picture books. The Touch could not replicate those very well, But the Tablet might. And math books are not there yet. But MathML is part of EPUB3 (or EPUB 3). EPUB is the core of the ebook creation and everything follows from EPUB. EPUB3 is really a subset of HTML5 (HTML5/CSS3/JS to be technical), so you might say EPUB3/HTML5 is the future of publishing.

Will people still want books? As time goes on, this desire will diminish. The reading devices are that good now. I suppose if someone wanted a book version of an ebook they would (via some program or service) convert the EPUB to PDF and send it to a Print-On-Demand (POD) machine to spit it out.

I don't know what will happen to the store with books. I don't think it has many years left. A big part of the bookstore itself isn't even books anymore. It's toys and other non-book gift items. And there is a section just selling NOOKs, of course. Maybe there will still be a store without books in the future where people go just to browse ebooks.

How to publish your own magazine app

... with Google Currents producer.

Caleb Hailey on Google+ shows the way:

Google Currents | A Game Changer?

See also:

Getting custom with Google Currents

from ClarkLab.


Here is my poetry blog published as a "current":

    google.com/producer/editions/CAowhLk3/poetical_bits


Note: If you have a Blogger blog and want to publish it as a Google current, it must first be "verified" that you are the owner. See:

    Verifying a Blogger blog in Webmaster Tools



For Flipboard, as for Currents, you could add a blog to your Google Reader (which is available to both Flipboard and Currents) and subscribe to there. With Currents, you can do more on the publishing side (via producer) than the RSS feed. Plus it runs on Android too.


2012/01/15: Tips to produce your blog to Google Currents (video)

Don't make sour ISBNade

When you think about this, it^* doesn't make much sense. It says that the EPUB and MOBI versions should have a different ISBN. But EPUB and MOBI are mostly interconvertable (it's basically a reformatting -- the same HTML and CSS is there) and do the same things (EPUB is a superset, really). Suppose you have an Android tablet and you get an EPUB ebook. In the software of the tablet one could have an embedded converter — like calibre calibre-ebook.com — that automatically converts the EPUB to a MOBI and you could be reading the ebook in a MOBI reader on that device vs. an EPUB reader on the same device that would look the same. Would you expect to see the MOBI ISBN on the copyright page or the EPUB ISBN? Only a fairly simple software transformation has been made.

Also I just got an ebook from Smashword smashwords.com and I got them in a bundle of formats: MOBI, EPUB, PDF, RTF, PDB, LRF, ... . What this recommendation says is that each one gets a different ISBN. (This was not the case, of course. It was the same ISBN on both the EPUB and MOBI versions, just as you would expect.) If someone sells their ebook in a bundle of formats like on Smashwords then there should be a different ISBN on each format? I don't think so.

I think the confusion is that the electronic world is not the print world. Just because there are different ISBNs on hardback and paperback (and those two from ebook), that analogy doesn't go over into the e-world since different e-formats can be automatically transformed from one to another, possibly in the reading device itself.

So the recommendation of different ISBNs for EPUB and MOBI is a bad one.


^* Re the Book Industry Study Group (BISG) proposal for ISBNs and ebooks, via the post When Book Publishing Gives You ISBNs, Make ISBNade by Thad McIlroy.


What I think will happen in practice for an e-publisher who is EPUB-based is that they would be able to sell their ebook (.epub) in the Nook Store, iBookstore, Google Bookstore, etc. with the ISBN (eISBN) they assigned to that ebook, but when it came to putting it in the Kindle store would let Amazon put an ASIN on it (as Amazon continues to be outside the EPUB standard).

From ADE to ereader

Here is the best researched guide I've found to the question of how to copy to your ereader an ebook you may have purchased from an ebookstore — represented as an ACSM ("Adobe Content Server Manager") file — that ultimately appears in Adobe Digital Editions on your computer:

How To Transfer Files from Adobe Editions to iPad or Kindle

This guide is written for PDF ebooks (and target devices Kindle and iPad), but the same steps can be followed for EPUB ebooks (and target devices NOOK, et al.).