Why Publishers Are Making a Push for EPUB3 Now

Expert publishing blog opinions are solely those of the blogger and not necessarily endorsed by DBW.

The Association of American Publishers (AAP) announced a major initiative to accelerate adoption of EPUB3 earlier this week. This latest version of EPUB, the widely used open standard format for digital publications, is based on the latest Web Standards including HTML5. But given that the vast majority of ebooks selling today are fiction and text-centric nonfiction titles that don’t have a clear need for the multimedia and interactivity associated with HTML5, some people have questioned why there’s such a fuss about moving to the new version.

The explanation is simple: While the long term benefits of moving up to HTML5 and the modern Web platform are significant as readers will demand more interactivity in ebooks, in the short-term the migration to EPUB3 is much less about ebooks that sing and dance and much more about lowering the costs of production and delivery for publishers across the increasingly wide range of devices, tablets and smartphones that readers are using.

To understand this it’s helpful to consider what version 3 adds to the EPUB standard. EPUB3 can be viewed as having four major feature clusters:

1. Styling and layout enhancements (general CSS enhancements, enhanced font/typography support, and fixed-layout support),

2. Global language support (vertical writing, R-L page progression direction, phonetic annotation, etc.)

3. Rich media and interactivity (audio, video, scripting)

4. Accessibility features (better semantics, pronunciation hints, synchronizing pre-recorded media with text display, mathematics etc.).

Many people think of HTML5 (and thus EPUB3 which differs from EPUB2 primarily in being built on HTML5) as being all about the rich media and interactivity. But HTML5 is a marketing term and in fact all of these enhancements in EPUB3 are enabled at least in part by HTML5 and related specs like CSS3 that come along with modern browsers. In some areas EPUB3 has gone beyond the browser baseline, such as in enabling global language support and accessibility. IDPF, the developers of EPUB, are collaborating with W3C to over time elevate the broader Web platform (which necessarily evolves a bit more slowly since it needs to address the needs of the entire global information and communications technology community not just book publishing).

Ebooks being sold today in the United States have been straight-jacketed to straight text thanks to the limitations of e-ink devices that have been prevalent in the last five years (and are only now starting to fade away as digital readers increasingly use tablets and smartphones). But it’s not clear that even on tablets that novels and narrative nonfiction, which form the bulk of the sales for most trade publishers, will benefit from video or interactivity. So that key feature of HTML5 is definitely not driving the AAP and other stakeholders to push EPUB3 adoption. Nor is it about global language support — in Japan, where EPUB3 is already widely used, vertical writing and related features have been critical but for English-language publishing these features are not necessary.

Instead, it’s the other two features of EPUB3 that are driving U.S. adoption. First, styling and layout enhancements. EPUB2 supported only a very minimal subset of CSS circa a dozen years ago and even for, say, a reflowable romance novel, the lack of reliable styling and embedded font support has been a challenge. Secondly, accessibility. Accessibility mandates apply of course to educational content, but also to other kinds of books used in government and educational settings; EPUB3’s baseline of accessibility support will eliminate the need to produce a second digital version using a specialized format like DAISY DTBook. Making content accessible may increase sales and of course is the right thing to do, but in many cases it is simply a cost of doing business that EPUB3 will reduce (and in any case work on DAISY has ceased and accessibility mandates are expected to shift to EPUB3).

The overall benefit really more about streamlining the workflow and supply chain so that a publisher can produce and reliably deliver one well-styled and accessible title as a single EPUB file to all distribution channels. We are mid-stream in the EPUB3 migration — with iBooks, Kobo, Google Play Books, VitalSource Bookshelf, and Sony having significant EPUB 3 support — and other distribution channels, including Amazon via KindleGen, supporting some EPUB3 features. But with many retailers and reading app developers still stuck on EPUB2, in the short term life has gotten worse for publishers, not better. So, making sure that over the next six months we get across the river to a place where EPUB3 is universally supported will make life much easier for publishers from a production distribution and quality assurance perspective, while in the process delivering the benefits of enhanced layout and accessibility.

But that’s just about optimizing what we are already doing successfully: selling novels and other text-centric ebooks. To expand the market and readership, we have to enable content that is being sold today in print form that isn’t selling well in ebook form via the older and more limited EPUB2 format. This includes e-textbooks, comics/manga, how-to books, children’s books, and magazines. Much of the need here is pretty basic: the fixed-layout part of the styling and layout enhancements.

For e-textbooks it’s going to be de rigeur to have integrated assessments, video, rotating 3D models, graphs of equations that are “live,” and other features that will require other parts of HTML5. Manga and comics are evolving towards being “motion books.” While the jury’s out on how much interactivity will be needed in what segments, I think within the next five years an ebook form of a history title will be expected to have slide shows and video as a matter of course.

Stepping back even further, what EPUB is really about is next-generation portable documents based on the Open Web Platform (aka HTML5). The publishing industry is moving inexorably to the Web platform as its core content delivery architecture: there is really no other viable choice. EPUB2 was a limited fork of the Web Platform circa 2001 and doesn’t get us there. PDF which is circa 1993 definitely doesn’t get us there. EPUB3 retargets EPUB to be based on the current Web Platform — not a fork but really built on the overall platform. And, EPUB will stay current (we are now prepping EPUB 3.0.1), and get better and better aligned over time, while helping to push the broader platform ahead with regards to publisher requirements. Once this transition is complete, publishers will have a lot more options for distribution including of course directly via their own websites and, where appropriate, as native apps. Of course you don’t need EPUB to develop a website or build a native app — EPUB simply provides a consistent framework for reliably structuring Web content so that it can be delivered in all of these ways depending on the circumstances.

It’s a very painful thing to leap forward a dozen years in one revision, and there’s definitely been some significant pain points in the migration to EPUB3 that’s been underway for well over a year now (and is taking longer and proving more challenging than many of us had anticipated). Kanter’s Law is that “Everything looks like a failure in the middle. Everyone loves inspiring beginnings and happy endings, it’s just the middles that involve hard work.” And right now we are surely in the painful middle with the migration to EPUB3 (as well as with the bigger migration from e-ink to tablets and smartphones and the even bigger overall migration from a print-only publishing industry to a world where digital forms of all types of content are expected).

The requisite hard work is well underway, and thanks to the momentum of the new AAP initiative, the Readium Foundation launched earlier this year, and other efforts by the publishing community around the world, I’m confident that by this time next year we will look back and all be very glad we’ve gotten across the river and fully adopted EPUB3 and integrated digital publishing with the modern Web Platform. And then we can get on with the harder and more interesting work of truly reinventing books and other publications — and publishing as a business — for the digital world. After all, EPUB3 and HTML5 are just the plumbing. In digital just as in print — maybe even more so — content is still king.

Read More About EPUB3:

— EPUB3: One File, Many Ebooks

— Evolution of Standards and Formats: The Costs and Benefits of EPUB3 and HTML5 Adoption

— Understanding EPUB 3.0 OnDemand WebCast Series

A Different View: HTML5 — The Future and Now of Publishing


11 thoughts on “Why Publishers Are Making a Push for EPUB3 Now

  1. Pingback: 趕快採用EPUB3標準吧Readmoo部落格 | Readmoo部落格

  2. Pingback: Why Publishers Are Making a Push for EPUB3 Now ...

  3. Pingback: On ePub3 adoptions | Sayegh e-Projects Blog

  4. Pingback: On ePub3 adoptions | Sayegh e-Projects Blog

  5. Pingback: ePub 3 | SES@: Eğitim Teknolojilerinin Sesi

  6. Pingback: Breaders Blog

  7. frank

    I unfortunately found out about EPUB3 when I recently purchased a book from Kobo and was unable to download the book into my computer and transfer it to my Sony non wifi reader. So now we have a push from the publishers for EPUB3 format that makes my Sony reader useless! I am pushing back, I will not buy an EPUB3 book and thanks to Kobo they gave me credit for the book I could not uses in my Reader.

  8. abhishek

    I was building an impression that epub3 is useless, and idpf peoples are forcing this standard on industry, as we know setting up the html5 standard wont automatically enable epub3 in most ebook readers and apps, but developers have to implement them painfully. To some extent idpf people assume that ebook reading system is standardized as and everyone is using webkit based rendering engine, which is not the case, if that is then every ebook reader would be 95% same. Overall I accept that epub3 is the future as you rightly pointed it would eliminate the needs of lot of other formats like PDF but there is a limit to epub3 usefulness and we shouldnot be doing excessive fancy things with it.

  9. Scott Miller

    Two years ago. Wow.
    I believe ePub 3 is critical technology, but even with it there’s still a missing piece.
    The 4th from the bottom paragraph talks about textbooks. File formats with gee wiz features are nice for textbooks (and technical books, instruction manuals, cookbooks, etc too), but they are not enough. The reason novels and other narrative-format books are popular on e-readers, but most other types of books are not, is the same reason novels and other narrative-format books are useful as audio books and other books are not. Many types of nonfiction are consumed in a nonlinear fashion. E-readers completely suck at allowing readers access an e-book’s content that way. Imagine textbook in audio book form–useless. And so too is at textbook on an e-reader. For nonlinear access traditional books blow away e-readers. They are faster and more intuitive to use.
    The Washington Post has an article on college students’ preference for traditional books over e-books. The focus is textbooks. Only 13% buy e-books or illegally download e-books from file sharing sites. The other 87% buy new, used, or rent traditional books. See: https://www.washingtonpost.com/local/why-digital-natives-prefer-reading-in-print-yes-you-read-that-right/2015/02/22/8596ca86-b871-11e4-9423-f3d0a1ec335c_story.html
    It’s a sick joke that computer technology (which is what e-readers are based on) should be so badly constrained by human designers and managers who completely lack vision. There are other computer-centric ways to view nonlinear information that blows away any e-reader device or app presently in existence. Imagine if Wikipedia was on your Kindle as an e-book. Are you going to start with the first article and read through to the last? You’d scream and cry, then gouge your eyes out in frustration. As presently implemented, e-readers suck. They are horrible devices.
    The technology is there, it just needs to be ported and tweaked for use as an e-reader.

    1. Mike Feury

      Tom, you miss the obvious, which was also mentioned in the article.

      Ereaders are mainly for text fiction enthusiasts. Tablets are for text and also other kinds of publications. Wikipedia works just fine on them and phones, no need to gouge any eyes out!

      Happy reading.



Your email address will not be published. Required fields are marked *