Notes from Editech (Italian EPUB Day) 2014

RSS  •  Permalink  •  Created 28 Sep 2014  •  Written by Alberto Pettarin

In this post I would like to collect some of my thoughts on what was presented and discussed at Editech 2014.

Venue, Programme, and Attendees

The "Italian EPUB Day (and a Half) 2014" was hosted in the H-Farm campus on September 25-26.

The programme included plenary sessions on THU 09/25, while on FRI 09/26 we had discussion tables and a final round-up plenary session.

The Thursday talks were the following:

  • Digital Publishing Standards: The Practical and Strategic Impact, by Bill McCoy, Executive Director IDPF
  • EPUB and EDUPUB: ongoing and upcoming developments, by Markus Gylling, CTO IDPF and Daisy Consortium
  • Digital Distribution, DRM & Open Source projects, by Hadrien Gardeur, CEO Feedbooks
  • Using EPUB3/EDUPUB within an LMS-mediated learning and teaching approach, Colin Smythe, IMS Global

plus, right before dinner, Riccardo Donadon (founder of H-Farm) gave a talk entitled Scenario e evoluzione dei modelli di business nel digitale (Digital business models: current scenarios and their evolution).

On Friday, each attendee sat on a discussion table in the morning, and on another in the afternoon. Each table was led by one of the speakers:

  1. EPUB 3 and EDUPUB: state of the art and evolution (M. Gylling)
  2. Digital distribution and DRM (H. Gardeur)
  3. EPUB 3 compliant reading solutions (B. McCoy)
  4. Production tools: EPUB 3 e-book creation (?)
  5. Metadata and discoverability (G. Bell)
  6. Using EPUB 3/EDUPUB within an LMS-mediated learning and teaching approach (C. Smythe)

and featured a "mentor", in charge of moderating the discussion and keeping it within the allocated time slot.

The work at each table was roughly split into:

  • 1 hour of discussion about the current scenario and future developments,
  • 1 hour of discussion about practical issues and Q&As with the table leader, and
  • 1 hour for summing up the discussion by producing a few slides to be presented to the other attendees during the final plenary session.

There were roughly 60 attendees, I would say 80% male in the 35-50 range. Almost everybody, except the speakers, from Italy.

A substantial majority of the attendees were direct employees working in publishing companies, either technical personnel of digital departments or managers of digital products. A few external services, software houses, and independent professionals were also represented.

I will not spend words on Thursday presentations, because they were pretty generic, while I found way more interesting the issues discussed on Friday during the working tables.

I only mention the fact that IDPF put online this Web page, detailing the status of ongoing specifications.

Morning Table: EPUB 3 and EDUPUB (led by Markus Gylling)

First of all, a much needed (for non-tech folks) clarification on the relationship between EPUB 3 and EDUPUB, which boils down to:

An EDUPUB file is an EPUB 3 file with constrained structure/metadata.

Technically, EDUPUB is a profile of EPUB 3. There is no such thing as a "EDUPUB format". This also means:

A reading system capable of rendering (all the features of) EPUB 3 files will also render EDUPUB files.

Then the discussion quickly headed to listing issues and desiderata for creating better eBooks, especially with respect to the new specs about indexes, dictionaries/glossaries, multiple renditions, and annotations.


The EPUB Indexes spec is currently a proposed specification.

We did not actually discussed about it that much, I guess because its obvious practical value, and its "nearly ready" status.


The EPUB Dictionaries and Glossaries spec is currently a working draft.

There is no public implementation (i.e., no app supporting it) yet, but Readium plans to support it by mid 2015.

After a cursory read, it looks quite good and reasonably well written. It allows authoring a dictionary (mono-, bi-, multi- lingual) as an EPUB file. It will also allow embedding a dictionary/glossary as part of a "regular" EPUB file, a feature that will surely prove interesting for e.g. complex novels (list of characters, places, artificial languages, etc.), non-fiction books, and of course textbooks.

I will probably write a future post on some issues with dictionaries, especially the integration within the reading system (e.g., stemming) and the fact that most dictionary indexes are stored in "binary" form (e.g., as a trie) to allow fast match and retrieval.

Multiple Renditions

The EPUB Multiple Renditions spec was probably the most discussed one, due to the high interest of publishers to distribute eBooks containing both a fixed layout rendition (mimicking the print version) of the publication, and a corresponding reflowable rendition.

Currently in draft status, it will also allow the creation of multilingual eBooks, suitable for distribution in different countries, and graded publications (e.g., "normal", "simplified", "expert", etc.) which might prove useful for students with special needs.

A rendition mapping document will allow the author to define the correspondence between locations in two (or more) renditions, thus allowing a reading system to switch from a rendition to another, without loosing the current reading location.

A nice corollary is that this specification will allow a declarative approach to the creation of parallel texts eBooks, as I have profusely written about already.


One of the arguments against current eBooks is the inability of the user to consistently create, store, transfer, share, and extract annotations and bookmarks.

Indeed, so far many reading solutions vendors have chosen to use proprietary, non-interoperable formats, resulting in significant limitations on the true potential of eBooks.

The Open Annotation in EPUB spec, currently in draft status, defines constraints on the W3C Open Annotation Data Model community draft to adapt it to EPUB publications.

Unfortunately based on CFIs (but IDPF might drop it in "EPUB 4"), Open Annotation will allow a reader to attach notes and bookmarks to an EPUB publication, to the extent of what the HTML5 object model allows: textual notes, handwritten notes/sketches (SVG), audio recordings, picture notes (images), etc.


So, these new specifications seem to allow for awesome eBooks, right?

Well, the real question here is if and when reading systems will support all these goodies. (See below for more on the topic.)

Additionally, the table also discussed about widgets, that is, embedding, inside an EPUB 3 eBook, Javascript code or other types of behavioral interactivity (e.g., 3D models). The idea that the creation of an ecosystem/marketplace of widgets and code snippets might prove commercially viable is surely interesting.

Finally, the table also produced some feedback on desiderata, e.g. distributing low-res assets while providing pointers to remote higher-res assets, which were particularly important for the publishers attending the session.

Afternoon Table: Reading Systems (led by Bill McCoy)

In the afternoon I sat on the "EPUB 3 compliant reading solutions" table, lead by Bill McCoy.

Here I brought my strong belief that current reading systems (including "mine") are quite poor with respect to what they can be, and this is a reason (maybe the reason) why eBooks have not taken off yet, as they do not offer a good experience to the user, in some cases even disrupting the eBook design of the author, thus crippling the investment of the publisher into authoring "good" eBooks.

Besides this general concern, publishers also appear to worry a lot about the consistency of the user experience: the same eBook, loaded on different apps, looks different. Even if I can understand the reasons for this concern, I do not care about it too much. The reason is simple: eventually, reading system will all offer two options to the reader, with respect to the "rendering" of the contents of an eBook: either stick to the publisher defaults, or apply customized settings (hopefully in a clearer way than tweaking the CSS as we all do today). And, of course, this will necessarily produce "formal/graphical inconsistency". What the publishers should really worry about is "functional inconsistency", that is, the fact that certain (basic) assets might not be consistently displayable/usable on each reading system.

This led to the request to IDPF/Readium Foundation to accelerate the development of the Readium project. In particular, the publishers would like to see the SDK evolve into a full, complete, polished app that they can suggest to their end customers, the readers.

Moreover, I also added that turning the Readium SDK into a FOSS project and eliminating the requirement to contribute in cash or code to get a commercially friendly license for it, would surely propel and improve both the Readium project itself and the EPUB ecosystem. Not to mention the "philosophical" implications of such a choice.

A list of practical suggestions to the Readium team was produced, including a better management of code and localization contributions, and a list of feature requests, ranging from supporting plugins to debugging tools, from annotations to better documentation and project communication.

I note that some publisher asked for a program of independent certification of reading system compliance to the EPUB 3 specification, especially for the EDUPUB profile, beyond current EPUBTest effort which runs on voluntarily (and often self-) reports. A final suggestion was to link epubcheck with the EPUBTest results, so that a publisher could immediately know about potential rendering problems on a given platform/reader.

On a parallel road, lobbying W3C for native support of EPUB inside the Web engines was cited as a potential action to increase EPUB reach and consistency.

Someone also proposed, as a temporary workaround for the "lack of good RS" issue, to facilitate the integration of Readium into app compilers like Cordova, so that an EPUB file (+ ReadiumJS) can be compiled into a native app for each of the major OSes on the market.

The Good, The Bad, and The Ugly

A few more notes about the format of the event and some personal opinions.

The Good

  • I was left with a sub-conscious feeling that if (and I admit it is a huge "if") apps are going to support all the new features (indexes, glossaries, dictionaries, annotations), then the future for the EPUB format might eventually turn rosier than its past.
  • The "discussion table" format promoted discussion of both short-term and long-term issues, providing bidirectional feedback between IDPF and attendees. (Let's hope that that feedback will be put into action soon.)
  • I met several technically competent people (more than I expected), passionate about the huge, exciting opportunities that eBooks might create.

The Bad

  • Those competent people sounded quite depressed by the current scenario, both in terms of spec/RS support, and the sustainability/ROI of advanced digital projects. (This tweet summarizes the feeling well.) And yes, I put myself in that club.
  • Apparently, not all table mentors were adequate. While the colleague moderating my table in the morning did a great job, the guy of my second table was evidently forced to sit there against his will. Besides not contributing to the discussion, he even left the table "for a quick Skype call"... lasted roughly one hour.
  • The room of the plenary talks: uncomfortable chairs, bad acoustics, no clip-mic.

The Ugly

  • I had the unpleasant impression, listening to the final presentations from the groups, that the discussion of certain tables was "suitably" guided by the speaker at the table.
  • Presentations in Italian at an event where nearly all speakers are from abroad? Really? I felt ashamed for the poor Indian guy sitting next to me, checking email and browsing during an 1h30m presentation in Italian.
  • Not surprising, and partially related: decent English verbal skills seem to be a quite rare among the Italian attendees. At the dinner, I was sitting in a table of eight, one being one of the foreign speakers. Since the other guys started to talk in Italian (rude!), I had the rather lucky opportunity to chat one-on-one with the speaker about metadata, accessibility, and PhD field memories.