In my consulting practice, all the book publishers I work with—whether trade, scholarly, or educational—routinely provide ePubs. It took a while for everybody to get on board with ePub as the e-book file format, but now it’s taken for granted.
But all too often those ePubs are based on an outdated spec. There are still a lot of ePub 2s out there—and too many are still being made. There’s no excuse for that except inertia: all the vendors know how to make ePub 3s, and it’s no more expensive to make a functionally equivalent ePub 3 compared to an ePub 2.
It’s likely that all the partners that make or receive a publisher’s ePubs not only accept ePub 3 but prefer it. Vendors only keep producing ePub 2s because publishers ask them to (or never bother to ask for ePub 3), and retailers and aggregators keep accepting ePub 2s because publishers keep sending them. (A famous book by Joseph Heller comes to mind.) Neither the vendors nor the recipients want to maintain two systems, two workflows.
EPub 3 is much better aligned with web technologies and W3C standards, so ePub 3s are more futureproof than ePub 2s. And ePub 3 is designed from the get-go to be made accessible. This isn’t new. ePub 3.0 came out in 2011 and was subsequently refined a bit as ePub 3.0.1, which didn’t make existing ePub 3.0s obsolete.
But a couple of years ago, when ePub was still governed by the IDPF, the ePub 3.1 spec was issued. It got almost nowhere. One reason was that, although ePub 3.1 was actually much clearer and better organized, and made some significant technical improvements, there were features in it that weren’t backward compatible with ePub 3.0/3.0.1. Understandably, recipients were reluctant to change their systems to support both versions. And if the recipients of ePubs didn’t accept ePub 3.1, publishers and their vendors weren’t going to make them.
Another obstacle was that ePubCheck, the software that ensures that an ePub is valid, was never updated to validate ePub 3.1s. Recipients of ePubs depend on ePubCheck. An ePub has to be valid or they won’t accept it—which means that publishers and their vendors also depend on ePubCheck. Not being able to validate ePub 3.1s was a deal breaker.
In the meantime, there was a very important development: the IDPF became part of the W3C, and the W3C took responsibility for ePub. At the time, there were those who worried that the W3C would neglect ePub or even drop it. The book publishing industry had become dependent on ePub. But those fears were unfounded. Work on ePub has gone on continually in the W3C since then. And a month ago, in May, the W3C issued the ePub 3.2 specification.
This is really significant for two reasons. It’s the first concrete development from the W3C on ePub. It demonstrates support not just for ePub but for the version of ePub on which the book publishing ecosystem depends. The work was done in what the W3C calls a Community Group, which is open to anybody and which costs nothing to join—you don’t need to be a member of the W3C to take part. That ensured that the folks working on ePub 3.2 came from a variety of types and sizes of publishers and other organizations.
Even more importantly, ePub 3.2 fixes the problems with ePub 3.1. And the best news: it’s actually not a big change at all. It will be really easy for anybody making ePub 3s to make ePub 3.2s. It won’t force recipients to reengineer their systems, and it won’t require publishers to update existing ePub 3s unless they have a reason to. Although ePub 3.2 incorporates many of the improvements in ePub 3.1, ePub 3.2 is backward compatible with ePub 3.0.1 (and thus ePub 3.0). That means all valid ePub 3.0s and 3.0.1s are valid ePub 3.2s.
Significantly, at the time ePub 3.2 was issued, ePubCheck 4.2.1 was released, which validates ePub 3.2 (and thus 3.0 and 3.0.1, too). The funds for that work were raised by the W3C Publishing Business Group, another component of Publishing@W3C that was created when the IDPF became part of the W3C.
So folks, the W3C has done well by publishing! We now have an excellent update to ePub that doesn’t make all our existing ePubs invalid, and we have an updated ePubCheck in sync with it.
Bill Kasdorf is principal at Kasdorf & Associates, a consultancy focusing on accessibility, information architecture, and editorial and production workflows. He is a founding partner of Publishing Technology Partners.
