EPUBJapan is a uniquely island nation, with a tendency to cling on to practices (the formal bow, the suit waistcoat, the command economy), long since abandoned elsewhere. And add to that list EPUB 3. For as it happens, Osamu Yoshiba, executive director for the digital and international business divisions of leading Japanese publishing house Kodansha Ltd., got up at Book Expo America to advocate its broad adoption as the answer to the general dearth of illustrated ebooks. (And there was me thinking it was something to do with display limitations.)

This Japanese attitude is easy enough to understand in Japanese terms. According to figures quoted by Digital Book World, around 80 percent of all ebooks published in Japan are Manga comics. Of these, around 80 percent are published in the visually friendly EPUB 3 format. No other major global publishing market has such a bias to accommodate.

To be fair, other publishing markets, including the Association of American Publishers, have also pushed for EPUB 3 in the past. But the problem is not with the publishers, it’s with the ereaders and the companies who support them – who are the de facto distributors. If you check out the E PUB 3 Support Grid over at EPUBTest, you’ll see that few major ereaders support more than 50 percent of EPUB 3’s features. Kindle manages about 30 percent, in Kindle or PC versions. Nook manages about 17 percent. Aldiko just clears 6 percent, and this is one of the most popular non-Kindle ereaders going.

Obviously with ereader makers so indifferent to EPUB 3, it’s hard to see the format as the basis for any new great leap forward in ebook publishing. But it wouldn’t be the first time that Japanese companies tried to ignore commercial realities.


  1. While it is literally true that EPUB3 support is lacking, I believe there is another part of the story not told here. The “major ereader makers” are really major ebook stores that make their money by locking customers into their platforms so they can sell books and make a hefty profit off each. The fact that they make an ereader too is just the price of being in that business. It is merely the delivery platform.

    These ebook sellers don’t have much motivation to adopt a public standard like EPUB 3. They have already developed their own publishing workflow, so they don’t care much about EPUB 3 publishing tools (not that they’ve really been developed fully yet anyway).

    There is a huge interest in EPUB 3 from textbook and technical publishers. Elsevier recently announced they are fully supporting EPUB and other publishers are following. The EDUPUB effort is also important in this area as it seeks to make an EPUB profile targeting education.

    While we might wish that EPUB interest would grow more quickly, it is definitely growing.

  2. It’s not really fair to say that “ereader makers [are] so indifferent to EPUB 3.” That’s a bit like saying automakers are indifferent to how well their cars are for transportation. The purpose of these readers is to read. It is that their ability to read ePub 3.0 varies enormously.

    The data tells a different story. Amazon’s main competitor, iBooks 3.2 for the iPad, offers 67.2% of all ePub features and 82% of the required ones. The latter is the best in the list. The equivalent figures for the iPad’s competitor, The Kindle Fire HD, is only 30.1% and 47.8%.

    Looking down the list, it’d be more accurate to say that Amazon’s support for ePub is lousy, that of many readers is poor, and that of Apple and Readium for Chrome (72.5% and 79.5%) is quite good. Those doing Manga cartoons can work with 80% compatibility (iPad). They can’t work well with only 30-50% compatibility (Kindle Fire).

    Also, keep in mind that what Osamu Yoshiba did, advocating ePub 3.0 for illustrated ebooks doesn’t imply a belief in the capabilities of current readers. In fact, his call suggests the very opposite, the need to make those readers better.

    Besides, Japanese publishers would only be technologically backwards if there were a competing format that’s newer and better and they were clinging to the ‘old’ epub 3.0. That’s not true.

    It’d be more accurate to criticize a backward-looking Amazon for clinging to an old format (mobi) and a proprietary one (KF8), when what’s needed it a common standard (ePub 3.0 and beyond). Leave the Japanese alone.

  3. The Readium Project includes a Javascript framework called Readium.js. As IDPF and it’s partners develop this framework, it will be possible to create eReaders that are fully HTML 5 and ePub 3 compliant. The web browser on your mobile or desktop device will be all that you need. It will be possible to store eBooks in the cloud or locally (using HTML 5 local storage) for offline reading or both.
    This is a very exciting prospect for those of us, such as in education, who are interested in dis-intermediating the publishing process so as to lower costs and increase access. The eBook could become the container of choice for teachers and students.

  4. Guys, the point of this is *not* to focus on Amazon, Apple, Nook, Sony, etc. – of course they’re likely to push for proprietary lock-in regardless of wider standards. It’s the independent ereader apps that interest me most. And of those, Aldiko sucks for EPUB3, as noted. Checking against the Feb 2014 ComputerWorld article below, the only other major Android ereader app in the article, Moon+ Reader, scores as badly as Aldiko, and the two others, FBReader and CoolReader, aren’t even on the grid.


    I use Mantano, which scores relatively well in EPUB3 compatibility. But I’d love to know the comparables for FBReader and CoolReader. Basically, if the big corporate monoliths diss your standard, that’s one anti-competitive thing, but if the independents do so as well, then perhaps you’re the one at fault.

  5. One of the problems with EPUB3 is that it is too loose a standard. It is somewhat difficult for publishers to produce an EPUB3 ebook and expect to work well on all EPUB3 reading systems. After all, that is really what everyone expects and wants.

    I believe the fault does not really lie with EPUB3 itself but with (a) the flexibility of HTML5, JavaScript, etc, (b) the tension between the desire for additional ebook functionality and maximum compatibility with reading systems, and (c) the immaturity of reading system implementations. Some of this will undoubtedly improve with time but there’s another way to tackle this problem.

    EDUPUB is a EPUB3 profile that seeks to create a standard within the EPUB3 standard that serves the needs of educational publishing. Their mantra is “Establishing a globally interoperable, accessible, open ecosystem for e-Textbooks and other Digital Learning Materials via EPUB 3 and the Open Web Platform”. Their hope is that both publishers and ebook reading systems will produce products that support EDUPUB and this will go a long way to solving the compatibility problems and provide an set of features that publishers and users can count on.

  6. Paul Topping said:
    “@Mike: That’s been done already. They are called ‘web pages.’”
    Not exactly. All of the web-based eReaders that I am aware of need an *.epub file which is actually a container for .xml, .html, .jpg and other file types. The best of these compete well on features with platform-specific eReader apps. The best eReader web app IMO was Ibis Reader. It made very good use of HTML 5. Unfortunately, it was taken down after O’Reilly Press acquired the developers, Threepress Consulting. It, too, required an unencumbered ePub file.
    The innards of ePub 3 files are increasingly dependent upon HTML 5, CSS 3 and Javascript, just as the WWW is but the container will probably persist. We could think of an eBook as a web site in a box – a box that can be conveniently transferred from one to another and function without a network.
    Still, I think that grappling with the question, “What’s the difference between an eBook and a web site?” is a great thought exercise.

  7. Web pages are on-line documents.
    Off-line html5 documents can replace ebooks and pdf. Ebooks just added a proprietary reader between the content and the reader. I suppose one can consider browsers proprietary but actually they are a helluva a lot more agnostic than eReaders. HTML5 docs can hold embedded media so the easy single file paradigm is maintained for many docs. And there is still the option for online media to be accessed from within the off-line document. One can also embed annotation technology. Really the sky is the limit.

  8. There is also an emerging trend in higher education whereby “learning content” is designed to be incorporated into a Learning Management System (LMS) in a way that obviates the need for any kind of textbook. The idea is that the institution purchases the learning content on behalf of the student incorporating the cost into tuition. This guarantees the publisher a 1:1 ratio between sales and enrollment. In return, the institution gets a lower per unit cost.
    Under he provisions of this scheme, there are no paper textbooks and there are no digital textbooks. There are only learning content modules that are inextricably tied to the LMS course. Since the LMS is a web-based system, these learning content modules are also web based but they are linked to LMS functions that enable “learning analytics” and other value adding features.
    Although this scheme would liberate the content in both paper and digital textbooks from their previous containers, that would only be to place it in another container, the LMS. It’s just switching jails and jailers.
    The anti-thesis of this is the trend toward using digital technologies to dis-intermediate educational publishers through self-published eTextbooks. That may sound very do-able but the resistance to change in higher education may well prevail.

  9. Frank: Thanks for pointing out that education (and anyone else, for that matter) have other choices for their content. However, you seem to imply that choosing EPUB involves some sort of publisher lock-in. There is nothing wrong with a teacher producing their own EPUBs and there are many tools already that do this. My company has been working with one called Scrivener from Literature and Latte that is an ebook authoring system. Apple’s iBooksAuthor also produces EPUB and is meant for non-professional use. I expect all word processors and page layout programs will have a Save as EPUB feature.

    Both the EPUB and EDUPUB works are open source and freely available. Though EDUPUB is being worked on by publishing companies targeting their needs, most of its features will be useful to all ebook productions.

  10. Paul, I certainly did not intend to imply that ePub involves publisher lock-in. Indeed, any enterprising teacher can create and disseminate ePub documents as you described. Educators do have the power to dis-intermediate educational publishing. The only question is whether they will make the (substantial) effort or not.
    However, educational publishers are certainly interested in vertical integration and the scheme I described will achieve that if widely implemented. Students will not go to the campus bookstore for their learning materials anymore. It will all come packaged in the LMS course. That scheme has nothing to do with ePub and, in fact, eschews it.

The TeleRead community values your civil and thoughtful comments. We use a cache, so expect a delay. Problems? E-mail newteleread@gmail.com.