Skip to main content

Bill McCoy is Wrong – Epub3 Isn’t Ready

4953147027_bee7cc6f7c[1]There’s a new article up on the PW website tonight, but I’m not sure it’s worth reading. Bill McCoy, the Executive Director of the IDPF, penned a post in which he tries to debunk 7 myths of digital publishing.

Unfortunately for Bill he managed to prove that one of his so-called myths isn’t a myth; it’s a reality.

In writing about Epub3 McCoy tries to argue that the Epub3 ebook format is ready, but the points he raises completely demolish his argument. Check it out:

Myth #4: EPub 3 Isn’t Ready

EPub 3 was rolled out as a standards specification in late 2011, promising support for rich interactive content and tighter integration of e-book standards with the full Web platform. Eighteen months later, while many reading systems support ePub 3, several prominent reading systems still support only the older ePub 2 standard. And even the vendors that do already support ePub 3 don’t support 100% of its features. How can publishers use a standard if it’s not uniformly supported across the industry?

There are two parts to resolving this dilemma. First, publishers can deploy ePub 3 content today that has enhancements that work on ePub 3 reading systems, but the content is also fully usable on ePub 2 reading systems or ePub 3 reading systems that lack some features. Every O’Reilly Media title published in 2013 is ePub 3, and in a Web post, the O’Reilly team explains how to structure content that is future-proofed as well as backwards-compatible. Secondly, realizing the need to “raise the bar” of full ePub 3 support ASAP, more than two dozen vendors and publishers have banded together to collaborate on an open-source ePub 3 implementation, forming last March the new Readium Foundation.

First, a minority of the ebook platforms that support Epub also support Epub3. It is misleading to use the word "many" to describe a minority.

And seriously, if you have to concede that it’s not fully deployed and that even the best support is incomplete, then you have basically proven that Epub3 is not ready. It’s not a myth; it’s a fact. And no amount of marketing hype will change that fact.

Update: It seems that even the IDPF has come to the conclusion that Epub3 isn’t ready. That’s why they announced in July 2013 that they’re going with an Epub3 Lite ebook format.

Just to be clear, I see Epub3 being held up not by the technical details of the format but by the lack of support among reading platforms. Apple supports most of Epub3, and Sony and Samsung support bits and pieces. So far as I know that’s it.

And that’s why I say Epub3 is not ready.

IMO, Epub3 won’t be ready for another 6 months to a year. Widespread adoption is waiting on the Readium SDK. This is going to be used as the core for reading apps and ereaders, and it is not expected to be ready until the end of 2013. Integrating the Readium SDK into apps and ereaders will take even longer (6 months to a year).

I know because I asked one of the developers only a couple weeks ago. I was double checking my facts for the post about why there are reasons not to adopt Epub3 just yet, and I was also told some other details that might interest you.

The reason Epub3 is waiting on Readium SDK is that the development costs are unbelievable (or so I was told). There’s a good reason why so many companies have joined the Readium Foundation:

I’d guess if you include dev, QE, design, product mgt, etc to create an EPUB3 renderer that is any good, and integrate it in to existing apps on all major platforms, you are looking at 10 yrs or something of man hours.

Yes, 10 years of labor.

Just to put that into perspective, let me remind you that we live in an age where a single developer can start a new service or company as a night job, where even successful startups can run with only a handful of employees.

With that in mind developing Epub3 from a technical spec into a usable ecosystem (app plus ebookstore plus distributor plus ereader) must be a huge undertaking.

Frankly, I am surprised that there has been any adoption of Epub3 at all.  Can you imagine how much Apple must have spent on developing the iBooks format, their own variant of Epub3? And then they followed up with Epub3 support. Of course Apple only had to support iBooks and Epub3 on IOS, and that made it a lot easier for them. But the expense involved still boggles my mind.

Folks, y’all are welcome to drink McCoy’s Kool-Aid, but I refuse. I plan to continue to ignore the marketing hype and look at the reality of the situation.

Epub3 is not ready yet. To say otherwise is to simply deny the facts in front of you.

P.S. Is it just me or is there something wrong when the only person speaking the truth is the outsider? That strikes me as a fundamental and frightening flaw in the system – any system.

image by RoxTues

Similar Articles


Tony Hursh May 29, 2013 um 12:06 am

IMO, EPUB3 will never be ready, in the sense of reaching large-scale deployment. It’s the X.400 of ebooks.

Ric Day May 29, 2013 um 2:00 am

I see this a little differently.

Amazon is unlikely to support EPUB3 because that would damage their platform lock-in strategy. They are far more likely to incorporate bits of it in their own rendering engine (at least for their more powerful tablets), but won’t be in a big hurry to do so.

Across the room, Apple supports enough of it already to make it interesting.

Sony is irrelevant; they barely exist as an ebook vendor. B&N sells a decent volume of eBooks in the US, but not elsewhere and their relevance may ultimately depend on Microsoft, who would probably want to compete in the lock-in space.

You didn’t mention Kobo – they have promised "full" EPUB3 support "by the third quarter" this year, which I take to mean they will have a working rendering engine on at least some of their reading devices by August or September and it will support a similar feature set to the one Apple is using. Kobo does sell decent numbers of ebooks.

Samsung? The Android crowd seem to mostly rely on 3rd party reading software. Bluefire is often a branded app on Android devices and appears to already support some elements of EPUB3.

So will EPUB3 become a "universal" standard? No, not a hope. The key players in the business wouldn’t let it happen. Each of them is engaged in as much customer lock-in as they can manage, so having their rendering software be a bit different is desirable.

What I think will happen is that sufficient elements of EPUB3 will be adopted that publishers will be able to produce ebooks which are not EPUB2.x that will work on most platforms. That is already starting to happen.

Readium is a different animal. Kobo and Bluefire and Sony are members; not so much Apple, Amazon, or B&N. I see the initiative as "interesting" but not yet compelling.

So think about what happened with (desktop) Web browsers – we endured decades of key players trying to own the reading experience. Dozens of browsers have been released and most have faded away. Today we have Chrome, FireFox, IE, Opera, and Safari. None is totally compatible with the others, despite W3C efforts to create "a standard."

I see the IPDF efforts as useful, to the extent there is at least a conversation around what we would like rendered in digital books. Bill McCoy’s comments in PW? Optimistic 😉

Nate Hoffelder May 29, 2013 um 7:48 am

The EPub3 efforts of Samsung and Sony are entirely independent of Readium/Bluefire, so they get an honorable mention at a minimum. Also, if they end up as being one of the few ebook platforms to support EPub3 then they won’t be quite so irrelevant, will they?

BTW, Bluefire is one of the core developers of Readium, and their support isn’t ready yet. Kobo is also a supporter, so I would bet that their EPub3 efforts are dependent on Readium.

Richard Pipe May 29, 2013 um 6:27 am

There are a lot more uses for ePub3 than e-retailer reading channels (although I grant that is a really big one). However the specification is written with so many loopholes there will be big enough differences in implementations to mean books will not be able to exploit all the features on all the platforms. Thing higher production and distribution costs!

We are delivering highly interactive ePub3 books to schools around the world with the AZARDI ePub3 reader and closed channel delivery. This is ePub3 on all platforms (Win, Mac, Linux) and devices (Android, IOS).

The biggest feature needed for education content, which neither iBooks nor Readium support correctly, is mixed fixed and reflowable. Here is our analysis of the situation put out just two days ago. Is ePub3 ready for Education . Other than iBooks it is the only commercial ePub3 reader, but we only supply it for institutional content delivery.

Bill has regularly responded at length to my criticisms of ePub3 and stopped mentioning AZARDI at his presentations even though we probably support it more completely than any other vendor.

There is a core set of features that must be supported in an ePub3 reader. The others are totally optional. We have a public analysis for this. There is also a kind-of more insanity going on with the current specification "advancements" being debated and documented.

In the AZARDI ePub3 reading systems we are about to start work adding support for the ePub Zero (E0) spec. That is going to be interesting.

fjtorres May 29, 2013 um 7:13 am

1- Readium is the reference design for epub3. Much as Adobe ADE for epub2.x, wherever the looseness of the spec produces conflicting interpretations, whatever Readium does will be the defacto "right" interpretation of the spec. Thus, if Readium isn’t ready, epub3 isn’t ready.

2- 10 man years development is is a *lot*? For an industry-defining core technology? Beg to differ. A 10-person team’s salaries for one year runs about US$1M. Maybe $1.5M.

Now, adding the hidden costs (payroll taxes, pension, dealth benefits) *will* double the cost So, go to $2-3M. Facilities and computers, support, Admin, etc? Hey, double it again. $4-6M.
That is a lot… if you’re a start-up in a garage.
But looking at the list of Readium "supporters", $5M, in absolute terms, is chump change. Adobe alone spends more just on janitorial. Or documentation for their creative suite. Note that Apple had no problem getting their own proprietrary implementation done in less than a year, and iBooks is a trivial part of the iTunes operation.

To put those numbers into perspective, a modern AAA-grade video game costs as much as $30M over 3-plus years to develop. That is $10M a year and it typically involves part-time efforts of as many as 300 people of literally hundreds of disciplines.
Of course, video games are expected to generate profits and the good ones do. The great ones generate gross returns in the ten-to-one range so their is incentive to invest is understandable.

$5M for a 10M-Y effort? I dunno, but that tells me that the corporate beancounters looked at the revenue projections and closed the wallets. There’s not enough money in epub3 any time soon for even a consortium of fairly deep pockets to justify the expense.

As I pointed out roundd hear a few years back, epub in all its flavors suffers from a commons problem; since it doesn’t belong to anybody, nobody is willing to sacrifice for it.

The epub3/Readium situation simply puts numbers on how much the major players in interoperable ebooks value their common spec and see its commercial prospects. No big prospects, no big investors. And vice-versa.

If even its deep-heeled supporters see little commercial value in investing in epub3…

Nate Hoffelder May 29, 2013 um 7:44 am

You might not think 10 years is a lot, but do remember that any number of Epub reading apps were developed by 1-person operations and released as noncommercial efforts. Calibre’s Epub support, for example drew mostly on Kovid’s work. And there were any number of reading apps on generic ereaders and tablets out of China. Few were any good, but they existed. There is even an Epub reading app for Windows CE.

Can the same be said for Epub3 reading apps? Not that I can tell. Sure, there a few Epub3 apps, but not as many as Epub achieved in 18 months.

fjtorres May 29, 2013 um 9:21 am

Your point is well taken and it’s another strike against epub3.

epub2.x took years to get established and *it* was a lightweight development effort suitable for startups.
epub3, however, is a corporate creature. A bloat of gold-plating and wishful thinking crafted by the marketting crowd. Without an opensource reference implementation like Readium it would take a master coder years to do on their own. And then the question becomes, why do it, no?
epub was needed to address broad market needs that individuals could commit to but epub3 isn’t targetted at broad market needs. Rather it is aimed at a collection of narrower niches and specialties that *might* develop into big markets. Those are corporate goals and in a corporate content ten man years is a modest effort.

As is, considering how "fast" Readium development is proceeding, I suspect the team may be working part-time on Readium and part-time on true line of business work.

Ric Day May 29, 2013 um 1:39 pm

@FJ complete agreement on the 10 years comment. The last large (publishing company) software project I worked on involved 110 people full-time for two years (that doesn’t include a couple of prior years of planning and hiring. So 220 person-years of actual software writing and testing. About US$25 million spent – back at the end of the 1980s, when t bought much more. Generates around $150~200m after tax annually, and dominates its specialized niche, so it paid off. 10 years is nothing – less than our planning time.

@Kevin – EPUB3 in certain instances does have a good use case argument right now (at least, parts do) and Apple is the one obvious target platform. I think that EPUB2 will dominate for a long time because it is "good enough" for plain, reflowable fiction, which means most novels. Anything that takes us away from the endless s in EPUB2 and lets us use more semantic markup has to be good!

I thing the initial appeal of EPUB3 will probably be to education publishers and consumer nonfiction, where the current fixed format layout possibilities in EPUB2 are a bit limited.

As a publisher, my immediate interests are far more skewed to workflow and tools in EPUB2, as I’m moving from the old print-first model to digital-first, InDesign and print (POD?) second. We produce ebooks in-house and finding fast and efficient workflows without paper is an interesting challenge. Having said that, we also have side experiments/testing going on to see how/if we can incorporate more of EPUB3 into ebooks and retain compatibility with EPUB2 (there was a good piece on this on the O’Reilly site recently). Metadata issues there are especially interesting.

I don’t think it was so much poor browser rendering as it was the desire of big multinationals to have strong DRM which drove the EPUB standard. You could render the pages of a print book in a scent CMS a long time ago, and create a TOC and Index.

Ebook publishing is likely to be a non-stop evolution for a few years more at least, much like any other software.

The more I think about this, the more I think EPUB3 will be a basket from which producers will pick the pieces they like. There will be a "drag" factor on consumer adoption simply because consumers who are happy with their current/recent reading device for the fiction they purchase in volume won’t have much interest in buying a new device anytime soon. Those old devices won’t be software upgraded to render EPUB3. And most publishers will be reluctant to produce multiple versions of ebooks for different devices.

Ric Day May 29, 2013 um 1:41 pm

Hmmm … my comment above about endless div’s in EPUB2 got munched by WordPress.

Ben May 29, 2013 um 8:28 am

Just stop talking about epub 3 already…shhhh. FJ has a great point, since it’s open source, no one is willing to sink too much of their own cash into i, only benefit others or their competitors using the same format.

Let’s not forget about publishers, most still outsource their ebooks for conversion. Which means any has little or not internal knowledge of how fast or slow the technology is changing. I often look at publishers' job listings, their requirements makes no sense or relevance for digital ebook related tasks.

Also don’t forget, bestselling ebooks are mostly fictions or text-heavy, all the interactive epub3 promises means little or nothing at all to publishers that are already doing well with epub2. They won’t invest in epub3 just because some dude at a book conference says it is the future of digital publishing.

I really hate some stuffs that comes out of book conferences, it’s often not based on reality and practicality. I’m glad Oreilly cancelled their conference. Stop feeding executives with things they don’t understand.

fjtorres May 29, 2013 um 9:42 am

It is easy to express support for a spec; send out a press release or trot out a talking drone and pull the string between the shoulder blades. Quick and cheap.

But providing tangible support?
That is where voting with the wallet counts.
Just how $trong their $upport is can ea$ily mea$ured, no?
Commitment to a corporate cause is easy to quantify and my take is that the epub3 supporters would *like* to see a mature epub3 but they don’t feel a *need* for it.
Not enough to, you know, actually provide measurable support.

Dave Bricker May 29, 2013 um 8:29 am

EBooks only exist because the web browser was originally unable to render HTML files in a flowable, paginated format. That’s all changed with HTML5. Javascript platforms like Monocle.js and TreeSaver.js have already solved that problem. With the majority of readers already consuming books on tablets and other non-dedicated-eReader devices, it makes more sense to return books to the web browser. The majority of mobile devices (plus all the desktop and laptop machines) have web browsers that support all of the ePub3 features that eReader manufacturers are dragging their feet on.

My hat is off to Bill McCoy and the ePub3 team but their innovations and contributions have been trampled on by a publishing industry that doesn’t want to adopt new standards they’ll have to design for. Converting a Word doc to ePub2 is one thing but designing a book to take advantage of the rich features ePub3 is another. Add concerns about alienating customers whose legacy devices will only support ePub2 books and you have an industry that sees change as an expense rather than as an opportunity to practice leadership.

Anyone who wants to produce an eBook that delivers on the vision of ePub3 has two choices: make an app and pay 30% to Apple, or build a web-based book and distribute as you would any other web-based content without the need for markup middlemen. Regrettably, though the ePub3 file format offers unlimited creative and expressive options, the devices that render it are not so visionary.

Kevin May 29, 2013 um 8:53 am

Nate and several commentators have talked about the iBooks proprietary format as though this is proof that EPUB 3 is never going to go anywhere. I just submitted an utterly compliant EPUB 3 textbook to Apple, who then took it and fiddled with it into whatever they do and put it into the iBookstore. I don’t care what they did. I’m just interested in giving them EPUB 3 files as the finished product.
While it’s perfectly true that we all want more supporting readers soon, any claim that EPUB 3 is unusable or dead in the water simply conflicts with my experience. People like me are using it to make really complicated multimedia books. If the claim is that we want more EPUB 3 support, I totally agree. If the claim is that EPUB 3 is unusable and always will be, then this is just patently false. I’m already using it to do things I never could before. The textbook has hundreds of illustrations and hundreds of sound files in it. This would simply not be possible without EPUB 3.

Ben May 29, 2013 um 10:09 am

I’m definitely not saying epub 3 is dead in the water. It is simply floating down the river instead of crossing it. ePub3 works on a much smaller scale (such as yourself) until the big players really makes the commitment with hardware and software support we simply won’t see a wide spread adaptation or the future it promises to deliver with ebooks.

I think Nate’s 1-yr prediction is too optimistic. These companies only care for their bottom line not how far they can advance technology. Happy that you can get your epub3 on Apple’s platform, does that translate to $$$? You can get back to us a year later. You also want to find out how well iBook Author ebooks are doing? (besides those textbooks?)

We are all enthusiastic interactive content or we wont be reading this stuff or doing it. Anyone remember interactive multimedia CD/DVD? Look at where that got us? I wouldn’t hold my breath and expect sudden breakthrough with epub3…yet.

Len Feldman May 29, 2013 um 1:49 pm

When the head of the IDPF, which is (nominally) responsible for EPUB3, has to name so many caveats and workarounds in order to just get EPUB3 content and readers to the point where they can reliably work with each other, it’s clear that EPUB3 isn’t ready for prime time. I lay the blame squarely at the feet of the IDPF itself:

* The IDPF knew from the outset that it would be difficult to write eReaders that could render everything in the EPUB3 standard correctly. However, it only recently defined a minimum feature subset that every EPUB3 eReader should support.

* The IDPF has no power to force anyone to comply with the full EPUB3 spec, a minimum feature subset, or anything in-between. I could call a can of tomato soup "EPUB3 compatible," and the IDPF couldn’t stop me.

* As we’ve already seen with EPUB 2.X, anyone can add proprietary extensions to EPUB, and not only won’t the IDPF stop them, it will defend their right to do so. I guarantee you that sooner or later, someone is going to add extensions to insure that their "EPUB3" eBooks only work on their eReaders.

These problems were evident from the beginning of EPUB3’s development, but no one in the IDPF’s leadership took them seriously.

fjtorres May 29, 2013 um 2:33 pm

They didn’t take them seriously? Maybe.
Or maybe they can’t get their members to commit enough resources to do anything meaningful.
Or, maybe they just don’t understand what setting a true technology standard involves.
IDPF seems to be more about politics and marketting than about standards management.

Drawing up specs is but the beginning of what is normally years and years of effort and *investment*. The faster you need it to go, the more expensive it gets. Trying to draft a workable standard on a shoestring, without involving any of the many world standards bodies is… ambitious. 😉

At *best* epub3 is headed where WiFi N trod; years of "Draft-N" products promising upgrades to the final spec. (Not all delivered.)
More likely, they are headed down the same road as SGML (with derivatives and partial implementations splintering the effort) or, worst of all, GOSIP (irrelevancy).

Joe Nonsense May 29, 2013 um 1:55 pm

LOL, you folks remind me of the late 80s naysayers against postscript based on holding up a page of Apple Laser Writer output to the light. Twenty-five years later show me where the photo typesetters are still in business. I am surprised at such a lack of imagination here!

The end goal is to move forward and away from all the static walled-garden formats and it will happen. Just as WordPerfect and Lotus 1-2-3, died they will too.

Len Feldman May 29, 2013 um 3:30 pm

Postscript was a language defined and controlled by one company (Adobe). If you wanted to claim that your printer was Postscript-compatible, you had to purchase a license from Adobe and certify that your printer was fully compatible. That insured that documents created with Adobe’s software (and software certified by Adobe) all printed out properly on printers that claimed Postscript compatibility. You could reverse-engineer Raster Image Processors that could interpret and render Postscript documents, but you couldn’t call them Postscript-compatible.

Compare that to EPUB3: Anyone can build an editor or eReader and call it EPUB3-compatible. There’s no requirement to validate it against the full EPUB3 specification in order to call it EPUB3-compatible. There’s no guarantee that an EPUB3 eBook from Vendor X will work in an EPUB3 eReader from Vendor Y.

For that matter, Apple, Barnes & Noble and Kobo all have their own proprietary extensions to EPUB 2 that support fixed-format layouts and interactivity. Those extensions don’t work in anyone else’s eReaders, even though the underlying files and file structure are nominally compatible with EPUB 2. Comparing Postscript with EPUB is nonsensical.

Joe Nonsense May 29, 2013 um 3:41 pm

LOL, you skipped 90% of the story.
But that’s not my point, which is about technology adoption over time.
How illogical is it to complain about incompatibilities of the different EPUB3 readers SO FAR to bolster the longevity of proprietary incompatible walled gardens of Apple, Amazon and others!
If you like, this discussion is similar to the browser wars. In the end the browser wars were lost, just like the epub wars will be defeated – by raw computing power which leaped over the petty walls. This also happened to Adobe after its Type 1 fiasco which led to TrueType.

Tony Hursh May 29, 2013 um 6:43 pm

" longevity of proprietary incompatible walled gardens"

You’re using Postscript as an example of "not a proprietary walled garden"?

"LOL" indeed.

Joe Nonsense May 30, 2013 um 4:07 am

My last point was directly to the contrary – Adobe’s walled garden led to the introduction of TrueType. While I dislike Adobe now, back in 1986 there was nothing cooler than PostScript. Hours after slamming my 75lb ALWP on the table, I had it connected my Mac’s 8 din to the printer’s rs232, fired up MacTerminal and after typing a few commands, the coolest pages were spitting out of the printer. Adobe sent me 1.6 mb of sample postscript code and I was able to write code to create my own brochure complete with graphic effects. There was no walled-garden when you are the only game in town.

While the naysayers harp on the weaknesses of epub3, there’s plenty of scope and plenty of time for improvement. And there is plenty of motivation – who wants to live in an Amazon-Apple dominated book world? And there is no logical reason to do so.

There are a lot of great basics in epub3 – html5, CSS2/3, SVG, Javascript, MathML, audio and video – a foundation from which to build.

And the basic requirements are simple – TOC-Manifest-Content. Each piece of content is a discrete piece of content and not constrained by other pieces of content. With technology like LESS, in a few years it will be bye bye to PDF.

PS to those dreaming of a perfect piece of technology, here’s a great example –

William Ockham May 29, 2013 um 4:45 pm

It’s depressing to watch a guy get beat up by his own strawman. Epub3 is an unmitigated disaster. It is one of the most poorly designed specs ever.

Bill McCoy May 30, 2013 um 6:14 am

"Ready" is a loaded term. By way of analogy: is HTML5 ready? Support for its features is incomplete and inconsistent across even the latest browserss, and MIA in the versions of IE with the most market share, the spec isn’t finalized, and HTML5 tools are not yet mature (native toolchains remain way more capable for mobile apps, and Flash still more the fastest way to implement interactivity). Facebook infamously decided to move away from HTML5 for its mobile apps. It’s pretty clear that Apple is unlikely to abandon either its proprietary Objective C based native app framework or its non-standard Dashboard widget framework, any more than Amazon is likely to abandon its non-standard KF8 format. So you could on the one hand credibly argue that "no, HTML5 is not ready".

On the other hand there’s hundreds of thousands of websites that use HTML5 features, with JS libraries that enable graceful degradation on older browsers, and a growing ecosystem of HTML-centric tools. Sencha demonstrated with their "Fastbook" experiment that Facebook’s reengineering was not based on hard limits of HTML5. And tons of people are successfully deploying cross-platform HTML5-based native mobile apps, using app-wrapping tools like Phonega. And while Flash isn’t quite dead yet,and full-native mobile app development is still thriving, the momentum is still shifting to HTML5 and the browser engine as the universal runtime. So in that sense, heck yeah "HTML5 is ready!" to use today, when and where it makes sense.

Similarly I see EPUB 3 as both "not ready" (especially in the sense of fully replacing EPUB 2 for use cases where that older version was sufficiently capable) and also "ready" (in the sense of already being commercially usable where it makes sense, including mechanisms to ensure content still works well with the previous version of EPUB systems). In Japan EPUB 2 was a non-starter (no vertical writing support), EPUB 3 is already the baseline format.

My article didn’t cover the duality of not-ready/ready and so probably came across as too optimistic. It was written as Seven "True Lies"… in the sense of conventional wisdom that had been or even still is true, but is in the process of becoming false. The editor suggested the change to "Myths" which is easier to understand but perhaps not quite as accurate.

Nate Hoffelder May 30, 2013 um 11:45 am

"in the sense of already being commercially usable where it makes sense, including mechanisms to ensure content still works well with the previous version of EPUB systems"

Is there a distributor ready to handle the complete spec? Is there an ebookstore that can support the complete spec? Other than Apple, that is.

If you can’t answer yes to both questions then in simple functional terms it is not commercially usable.

Uneducated Guess May 30, 2013 um 10:11 am

Right. So we have the web, and we have native apps. Now we need a new little EPUB3 to download that qualifies as either the world’s most disconnected website or the world’s lamest app. Why would I want this?

Eping Wang June 4, 2013 um 4:35 am

I think Adobe is very reluctant to develope more on ePub, which has a possibility of doing harm to its very profitable PDF.

EPUB 3: Der Stand der Dinge | digital publishing competence June 4, 2013 um 6:02 am

[…] muss er das natürlich). Demgegenüber weist Nate Hoffelder auf The Digital Reader zu Recht darauf hin, dass diese Aussage doch zumindest einige Realitäten des eBook-Marktes verkennt. Auf und im […]

EPUB 3: Warum der Standard auf sich warten lässt » eBooks, Software » June 4, 2013 um 8:41 am

[…] muss er das natürlich). Demgegenüber weist Nate Hoffelder auf The Digital Reader zu Recht darauf hin, dass diese Aussage doch zumindest einige Realitäten des eBook-Marktes verkennt. Auf und im […]

Nil Einne June 8, 2013 um 5:29 pm

One point you’re perhaps missing is multilingual support particularly for non latin alphabet languages with more complicated or at least different rendering. While there appear to be various unofficial ways to do vertical text in Epub2 and it sort of works in the Adobe rendering engine used by Kobo for sideloaded books, from my own experience and of others it’s rather problematic. Similarly while it appears to have some support for right to left text, it doesn’t support the logical and basic extension, right to left page progression. My test have been in Japanese, I haven’t found a good sample for traditional Chinese with such features but I presume it’s the same. I haven’t personally tested but I expect and from what I’ve read it’s true that the support for any sort of complex script/text rendering is limited. And again from what’ve read support isn’t much better in other ereaders using ePub2, in fact often worse.

Meanwhile the Kobo kepub reader/renderer used for Kobo books and sideloaded books with the kepub.epub extension works far better on these albeit still sometimes needing some fiddling. The kepub reader appears to be Kobo’s epub3 targetted reader with some support for fixed layout etc (far, far better than the Adobe one) and I guess is either going to surplant the Adobe ePub2 reader entirely or be replaced itself in the future when they official support ePub3. The kepub reader unlike the Adobe ePub reader also seems smart enough to use another font if the glpyh it needs is missing from the selected font although seems to have font size issues then.

So all this goes back to want point namely if you’re a publisher of works requiring these, your options are rather shitty but it seems clear targetting ePub3 is your best bet despite the limited support. You may want to also target KF8 but since that appears to be largely derived from ePub3 it’s no biggie adding that as well.

You may consider this a 'niche' market.

Tony Hursh June 8, 2013 um 6:39 pm

EPUB2 content is XHTML which supports both Chinese and Japanese.

If there’s a problem, it’s with Adobe’s rendering engine or other reader software, not EPUB2 itself.

MURATA Makoto July 19, 2013 um 7:58 pm

Actually, EPUB2 does not support vertical writing, ruby, and so forth.

Nil Einne June 8, 2013 um 5:31 pm

Sorry meant to say, you may consider this a 'niche' market, but it’s still a market and likely a growing one interested in digital publishing and ereaders.

Nathan June 11, 2013 um 4:29 pm

There are several reasons why EPUB3 is something that should be supported, however eBook distributors choose to come up with their own proprietary formats which sets back EPUB3. For one, fixed-layout. Most people don’t want to distribute to just Apple or Amazon (KF8). EPUB3 pretty much allows fixed-layout without actually being Apple’s fixed-layout. And if everyone would support EPUB3, it would require only one file. Also as stated above, languages. While Japanese and Chinese characters may be supported (Apple requires EPUB3 for some languages) the vertical layout or right-to-left layout isn’t.

Mers July 8, 2013 um 9:16 pm

To some ebook developers, creating ebooks to many formats starts with creating an EPUB file.

MURATA Makoto July 17, 2013 um 9:09 pm

I am the lead of the Enhanced Global Language Support sub-group of the IDPF EPUB WG.

In Japan, EPUB2 has not been used, but EPUB3 is now everywhere. Implementations are still unsatisfactory, but they are getting better and better. Legacy formats are shrinking. Conversion from EPUB to Kindle mobi is widely used as a vehicle for creating vertical-writing Kindle e-books.

The reason of the success of EPUB3 in Japan is the addition of vertical writing and other I18N features. I am confident that the same applies to Taiwan. I also know that Korea is committed to EPUB3. The support of BIDI (required for Hebrew and Arabic) in EPUB2 was not satisfactory, because it was not possible to indicate the right to left page progression.

The world-wide market of EPUB3 is very attractive for many players. I am confident that EPUB3 is the crucial e-book format for the world.

Mers July 20, 2013 um 12:15 am

This is correct. The world needs EPUB3 because of the many features it offers (Bi-directional and vertical text, video and audio support, MathML, etc.). Features that are readily available to many developers and that don’t need additional plug-ins or 3rd party software.

Device manufacturers are making the promise to support EPUB3 and we should be happy that they, on their part, are TRYING their best not seal ebooks on their own proprietary format which Amazon did. EPUB3 will be what like Liz Castro said on her "EPUB in the Wild" wherein EPUB3 is a file that can be read by many e-devices. A file not constrained to be read on a single device.

EPUB3 also gives rookie device manufacturers an edge in the market. If, for example, I plan to create my own ereader, creating my proprietary format which no one else supports yet will not help me break any market. But if I choose to have my own ereader file format AND support EPUB, then I will gather some share of consumers in the market since EPUB is not constrained to any single ereader.

Also, If I have an EPUB that I read on my iBooks/iPad, I’d be able to easily transfer it to any other ereader that can read EPUB. In some way, this can be called "FREEDOM" (not piracy) which is another feature that EPUB (be it 2 or 3) has to offer.

I read so many negative feedbacks about EPUB on the internet and it’s very disheartening. People who are not knowledgeable about EPUB are being misled by these ignorant posters and bloggers. These are people who don’t really know and understand EPUB. They keep complaining that "It lacks this and that". What is really lacking is the ereader that doesn’t support much of what EPUB has to offer. These are also the people that are selfish that they want their own certain feature to be implemented.

EPUB3 may not be ready, so they say. But it’s getting there. It has a vision and mission that is not selfish and controlling. EPUB is FREEDOM. Those who deny the goodness of EPUB are ignorant and those who say that it isn’t ready are people who are not willing to start anything new and exciting in their lives.

Tony Hursh July 20, 2013 um 12:28 am

"(Bi-directional and vertical text, video and audio support, MathML, etc.)."

None of those require epub3. They’re all in plain ol' HTML5.

Mers July 20, 2013 um 12:50 am


Tony Hursh July 20, 2013 um 10:14 am

No, any more than HTML5 is bits.

The problems with EPUB3 stem from the incredibly baroque and "overdesigned by committee" spec, not with the raw HTML underneath.

Mers July 20, 2013 um 1:06 am

Also Mr. Hursh, HTML5 is not plain ol'. It’s new…

MURATA Makoto October 4, 2013 um 8:03 pm

BIDI is certainly supported in HTML5 and even earlier version of HTML. However, to use BIDI for EPUB, we need a mechanism for specifying the page progression direction. This does not exist in EPUB2.

Vertical text is going to be provided by CSS Writing Modes. It would not have happened without EPUB3.

Mers July 20, 2013 um 12:51 am

I mean EPUB3 uses HTML5. Other proprietary format I don’t know what they use. Have you even opened an EPUB file Mr. Hursh?

Richard Pipe July 20, 2013 um 1:03 pm

I have to agree with Tony. I fact his statement "incredibly baroque" and "overdesigned by committee" hits THE nail on the head.

There is a complete lack of cohesion and direction in the ePub3 spec. It is like a number of "bright-boy" brain dumps. EPub3 is a rigid XML container format with XHTML5 content files with custom name-spaces. At best it is a sort-of HTML5. However that is not the problem. It is the yesterday reading system add-ons that really hurt future looking digital content publishing.

The good news is that the core packaging spec is reasonably OK, but still arcane XML. It reflects the yesterday specification writing background and approach of the authors. Like all such projects the authors forget the content and focus on "the Schema" and "the rules". It does little for publishers while significantly increasing cost of production and total cost of content ownership. It is written for reading system implementers.

There is a lack of coherence in the spec. There is a drive to define a rigidity that is counter to the requirements of digital content now. It doesn’t encapsulate the sophistication of HTML5. It murders it.

The result will be core package support with a diaspora of ad-hoc reading system implementations (like iBooks andReadium). Except for basic novels and simple fixed layout books there will be no significant implementation (nor should there be) of the "advanced features".

For example neither Readium nor iBooks support fixed layout spine properties because it was too hard… or not important. They are not even doing it with the Readium SDK because vertical reading for Japanese is far more important. Correct use of the fixed layout spec would also directly compete with iPublish, Inkling and Articulate (claimed HTML5 proprietary packages). These all natively use combination fixed layout and reflowable content for education content. The direct neglect of education content in any aspect of the specification or implementations is stunning.

No matter how brilliant an implementation of the specification the Readium SDK may become, it can’t take into account the cost of creating the content for the reading-system-centric features because the specification doesn’t. The ePub3 specification has just forgotten it is the publisher and the content that matter.

We are supports of ePub3. We create and maintain the AZARDI ePub3 reading system, have a sophisticated ePub3 production system and have delivered thousands of ePub3 education and learning titles. That doesn’t mean we can’t and shouldn’t criticise the specification text, tone, direction and capabilities.

EPub3 is a valid, but shaky, learning stepping stone as digital content matures. While it may have looked a little snappy in 2011, it just doesn’t deliver what is required in 2013. As a publisher format it is does nothing for linear reading content, is marginally better than ePub2 for some other types of content and is a failure for everything else; unless supported by a reader that matches the implemented content to reading system features.

Mers July 20, 2013 um 9:11 pm

Like what you said…

"… is a failure for everything else; unless supported by a reader that matches the implemented content to reading system features."

EPUB3 is not the problem. People who create ereaders like AZARDI have a problem on HOW to implement EPUB3 specifications. When they encounter the problem, they blame it all on the committee of IDPF, saying it lacks "cohesion and direction". Now when consumers encounter a problem with the ereader, the blame could be put on the EPUB committee rather the ereader itself. Sad thing is, what might be badly formatted in one ereader might be okay on another. That’s why EPUB developers have EPUB readers like ADE (which many say is badly developed), iBooks, Nook, etc., to check the EPUBs. Each ereader having their own interpretation on how to display a certain HTML tag. Those differences in ereader interpretation causes the problem, not the rules that have been implemented.

These problems will be the very things that ereader developers will encounter if they plan to jump in the ereader market. Ereaders are everywhere. But not one has full support of EPUB3. Or so may I say, not one has full support of HTML5 features. Because even the 5 major browsers (Chrome, Firefox, Safari, IE10, and Opera) don’t have full support of HTML5 (HTML5 is not final yet) which is the core of EPUB3. All these technologies are interrelated.

I found out about AZARDI when I looked up for an EPUB3 reader. I didn’t realize you are against a committee in which your ereader basically support the fundamentals. If we have a problem with EPUB3, then let the committee hear us. But let’s not denounce EPUB3 in its youth. Everything, and everyone, has flaws. The flaws will only be resolved if we speak to the committee about it. The HTML5 specs is not yet final, it is still a recommendation till 2014. HTML5 is the core of EPUB3 and if HTML5 is still a recommendation, what more of EPUB3? And HTML5.1 and 5.2 is also coming. Just think how that will affect EPUB in the long run. Just as HTML5 is young, so is EPUB3.

By the way, despite it all, I like AZARDI. A very well developed ereader for EPUB3.

Nate Hoffelder July 24, 2013 um 4:47 pm

It turns out that you are wrong. Epub3 is so difficult to implement that the IDPF is going to develop an Epub3 Lite format:

Mers July 25, 2013 um 2:54 am

According to your quoted material…

"The initiative’s goal is to accelerate the across-the-board adoption of the EPUB 3 format in the consumer market by identifying and implementing what stakeholders consider to be the core set of baseline features critical to the format’s acceptance. Among these features are greater interactivity for users, multimedia-enhanced content and expanded accessibility for people who are blind or have other print disabilities."

It didn’t say EPUB3 is not ready…

Nate Hoffelder July 25, 2013 um 7:23 am

If Epub3 were ready then there would be no need to accelerate implementation.

Mers July 25, 2013 um 8:23 am

You know what is preventing EPUB3 from accelerating? It’s people like you who thinks it’s not ready when it’s already here and happening. The fact that it needs to accelerate means it has already started. Jumpstart would have been the word if it hasn’t started yet, but the word is "accelerate". Why are you so against EPUB3? Is it because you want to prove that you were right in one of your previous post about the “dead on arrival” of EPUB3?

I could accept the fact that one of the reasons EPUB3 is "lagging" is because some walled-garden publishers want to implement their own proprietary formats. But for someone who doesn’t have his own proprietary format and complaining about EPUB3? Prove first that you can do something better than EPUB then start complaining. Otherwise, you’re just someone who predicts that EPUB will die and then when it happens, will say to the world "I told you so". You just want to prove your point.

You may be right, you may be wrong. But in the end, if you were right, that wouldn’t help you in anyway, and you wouldn’t need to tell anybody "I told you so". Because when EPUB dies, when it dies, many people might have already found another technology to replace it. Another form of technology that you would probably declare "dead on arrival" again.

Some publishers like O’Reilly have already embraced the technology despite the setbacks you’ve been scratching. And many others are making the promise. Don’t you want that? Don’t you want EPUB? For accessibility? For the blind and the print deficient readers? Don’t you want multi-language support? Don’t you want an ebook format that is not constrained to a single ereader?

No sir, you want to be right. You want to proclaim that you were right in declaring EPUB to be dead. Forget about the blind and the print deficient readers. Forget multi-language support. Forget EPUB. Then what? Let’s all go to Amazon, or Apple iBookstore, or any walled-garden publishers then support their format? A sure sign of division in the ebook format when what we actually need is one single file format that unites all ereaders. That is EPUB, a format that has a vision of unity despite the division among ereaders. But you don’t want that. The way you write against it goes to show.

I could rant all the way and you wouldn’t bother thinking about anything I’ve written. But I wrote these words anyway because I know that EPUB is a right thing and I want to fight for something that is right, even if it’s flawed. EPUB may die, it could die, because not everything last forever right? If it does, it will bury with it the things that could have been good and great for the publishing industry. The greatness that couldn’t be given by the one who predicted its death.

Nate Hoffelder July 25, 2013 um 8:39 am

"You know what is preventing EPUB3 from accelerating? It’s people like you who thinks it’s not ready when it’s already here and happening. "

So Epub3 is happening in Kobo and Nook? Not hardly. Little bits of Epub3 is supported by Kobo and that’s it.

And it is rather amusing that a gadfly can have such an effect on the technical implementation of a format. I didn’t get interested in why Epub3 2as delayed until early this year, and yet I still managed to delay adoption in 2012. I am skilled.

Where are the indie apps? The Epub3 spec was finalized 21 months ago. By the time the Epub spec had been around for a couple years there were a dozen different apps and devices that supported it, including some for niche platforms like Windows CE. You can’t say the same for Epub3, and that is a sign that it has stalled.

"I could accept the fact that one of the reasons EPUB3 is “lagging” is because some walled-garden publishers want to implement their own proprietary formats."

I didn’t say that.

"But for someone who doesn’t have his own proprietary format and complaining about EPUB3? Prove first that you can do something better than EPUB then start complaining."

LOL That’s like saying that because I’m not a car manufacturer I can’t point out that your wheels are going to fall of of your car. Anyone with eyes and a brain can see what is going to happen.

"For accessibility? For the blind and the print deficient readers? Don’t you want multi-language support? Don’t you want an ebook format that is not constrained to a single ereader?"

So you are saying that we won’t have those things without Epub3? That’s a false argument.

Mers July 25, 2013 um 8:50 am

About the wheels that would fall off the car, that’s fixable. Anyone with half a brain knows that. You think EPUB is fixable?

Then again I ask you, can you offer anything better than EPUB? That’s the most important thing that you’ve missed.

Nate Hoffelder July 25, 2013 um 9:23 am

Can I do the work personally? No, but that is irrelevant because it leads back to the false car manufacturer analogy.

If you want suggestions, I have 2. Epub3 could be trimmed back what Epub2.x devices now support (plus metadata).

Or Epub3 could be trimmed back to whatever Webkit can support out of the box (plus metadata). This second option had been proposed as the likely outcome as far back as early 2012. It has the benefit of having low marginal cost for implementation while still packing in features.

Mers July 25, 2013 um 10:09 am

The first option only takes us back to EPUB2.

I’d go for the second option… interesting option that needs a looking into…

Thanks for the answers. I very much appreciate it.

Oh My God July 25, 2013 um 12:20 pm

The real McCoy
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.”

Nate Hoffelder July 25, 2013 um 4:59 pm

This is the same guy that wrote back in May that Epub3 was ready, so you might want to take his words with a pound of salt.

Oh My God July 25, 2013 um 5:16 pm

Yeah who is that guy? Stuck in the middle? A nowhere man.

Nate Hoffelder July 25, 2013 um 5:40 pm


Bill McCoy July 25, 2013 um 8:24 pm

Well Nate, I’ve made the case that EPUB 3 will be broadly adopted and you’ve made the case that it will be a "dismal failure". In six months or less we’ll know which one of us was spewing BS and which one of us should be taken seriously.

Nate Hoffelder July 25, 2013 um 10:04 pm

Do you know what?

That was a stupid, rude, and inappropriate thing for me to say. I apologize.

EPUB 3: Der Stand der Dinge | digital publishing competence April 16, 2014 um 8:38 am

[…] muss er das natürlich). Demgegenüber weist Nate Hoffelder auf The Digital Reader zu Recht darauf hin, dass diese Aussage doch zumindest einige Realitäten des eBook-Marktes verkennt. Auf und im […]

L May 28, 2014 um 4:39 pm

I’m late to the discussion and I don’t know much about ebook formatting, but I ran into my first epub3 problem today. I purchased a book from kobo, but when I went to download the ePub, that option was not available. I contacted kobo to get my money back, they informed me the book was only available in epub3, that was why downloading was not an option. Although in theory I knew epub3 might not be readable on every device, I didn’t know it would mean I would be unable to even download a copy. I don’t like where this seems to be headed.

Adobe, You’re Making Me Squirm | MVO Productions September 26, 2014 um 4:06 pm

[…] ** Updated-9/26/14: For some more strongly worded critique of the EPUB3, read Bill McCoy’s post. […]

Write a Comment