Posts

Inclusive Publishing Hearts and Minds at the Digital Publishing Summit

DPUB logoThe 2019 edition of the Digital Publishing Summit took place in Paris (France) on June 25-26. This increasingly popular two-day conference—now in its fourth iteration—is organised by the not-for-profit organization EDRLab (European Digital Reading Lab). The conference began in 2016 in the town of Bordeaux (France), then moved to Brussels (Belgium), followed by Berlin (Germany).

This year’s opus was hosted by the National Library of France, in the capital city where EDRLab’s headquarters are located and has, yet again, been a showcase of some of the most exciting developments in the industry. The two conference days were packed with in-depth presentations as well as entertaining insights on the many facets of the publishing ecosystem. Business managers and technical experts shared their experiences, and reflected on the challenges and opportunities that have marked the year.

The extensive program covered a broad spectrum of topics, including:

  • The essential role of technical open standards—the EPUB specifications, the W3C web platform, and the global collaboration to normalize digital sequential art / visual narratives (comic books, manga)
  • The strategic significance of open source software—EPUBCheck and the Ace by DAISY Accessibility Checker
  • Innovative research and development projects coming to fruition and the growing adoption of Readium architecture in commercial-grade solutions
  • User-friendly Digital Rights Management gaining traction in the library context—Readium LCP (Licensed Content Protection)
  • The emergence of bleeding-edge software solutions and services, automated production processes powered by machine learning principles
  • Changing consumer habits and evolving business models including  the remarkable expansion of the audio-books market
  • Legal directives and legislative frameworks, ensuring inclusive access to all published content—the European Accessibility Act and the ongoing ratification and implementation of the Marrakesh Treaty

This short article only touches the surface of some of these topics, so make sure you explore the YouTube video recordings of the conference talks.

Daniel Weck presenting at DPUB with a power point slide containing the Thorium logo

Image supplied by Luc Audrain, Hachette Livre

I personally made a presentation about Thorium, an open-source desktop application based on the Readium architecture. Thorium is a reading system developed by EDRLab, which aims to support EPUB (reflowable and fixed layout), the LCP DRM, OPDS catalogs, TTS read aloud, accessible annotations, and many other essential features. I talked about usability, as well as the technical challenges faced when implementing accessibility techniques (e.g. keyboard navigation, screen reader support, customizable text formatting, adaptable user interface, etc.). Aferdita Muriqi (EDRLab) showcased the Readium mobile apps, with the notable addition of support for audio books. Naturally, accessibility is also high on the priority list of the Android and iOS projects.

Over the past few years, the digital publishing industry has demonstrated a passionate drive to create systems with inclusive design in mind, impacting not only the reading experience end of the supply chain, but throughout the entire ecosystem—authoring tools, production practices, validation utilities, delivery and discovery systems, and of course technical standards and publication formats.

Ken Jones' first slide at the #dpubsummit a11y=accessibility

In the first slide of his presentation, Ken Jones (Circular Software) showed how the word accessibility contracts to the term a11y. Ironically, Ken immediately went on to make live demonstrations of interactive, animated, all singing-and-dancing fixed layout EPUB publications, which usually rate low on the a11y-o-meter! However, Ken made the case that documents authored with Adobe’s InDesign—a tool typically geared towards visual typesetting and graphic-oriented creations—can indeed feature a clean navigation structure and linear reading order. Ken showed keyboard usage and gave an example of accessible text in typographically-rich layouts. He also explained how to label interactive controls so they are announced by screen readers as well as demonstrating EPUB Media Overlays read aloud— the playback of audio narration synchronized with highlighted text.

There is indeed a strong case to be made for optimized publications – as defined in the EPUB Accessibility specification, for example those that meet the particular needs of readers with dyslexia, using adapted typography, text formatting, colours,  and synchronized voice narration, etc.

However, fixed layout EPUBs have an intrinsic characteristic that makes them inherently less accessible than their reflowable counterparts: they are not authored with responsive design in mind. By definition, a fixed layout document is not meant to reflow depending on screen size and device orientation. At best, it can be scaled in or out, and panned across. Crucially, in the real world, many fixed layout publications are just mere exports from typesetting tools like InDesign, or straight conversions from PDF with little or no care for accessibility.

A case in point: Vincent Wartelle (ISICrunch) gave an entire presentation about using machine learning—more commonly referred-to as artificial intelligence—to reverse-engineer the navigation structure, linear reading order, and text flow from PDF files typically published in the education sector. In my opinion, although there are valid justifications for fixed layout EPUBs, I believe the first port of call for any modern digital creation should be a reflowable responsive design.

Thankfully, nowadays there are readily-available tools to help validate publications not only in terms of their correctness with respect to the EPUB specification (i.e. EPUBCheck), but also by asserting a number of well-defined accessibility criteria. As Romain Deltour (DAISY Consortium) demonstrated, the Ace by DAISY Accessibility checker verifies W3C WCAG rules as well as EPUB-specific ones. The report produced by Ace contains a list of categorized and prioritized issues, which content creators can use to identify problems and address them effectively (references to the DAISY Knowledge Base are provided).

@wendy_a_reid @kobo talks about audio books at #dpubsummitI anticipate that in some not-too-distant future, there will be similar validation/checking tools to help with the production of audio books. In her presentation, Wendy Reid (Kobo) walked us through the W3C audio books standardization effort, which includes a strong recognition of the need for a structured table of contents (not just a playlist of MP3 files!), accessible descriptions, as well as the potential of more advanced features like escapability and skippability (which are well-known concepts in DAISY Digital Talking Books)

Many of the presentations at the Digital Publishing Summit did not have a direct accessibility angle, which made them no less interesting. Julie Blanc (Labo Paragraph) talked about “page.js”, an HTML toolkit for print layout based on CSS stylesheets. Florian Dupas (Kwalia) presented the latest research and development in the field of digital comic books – sequential art / visual narrative – a collaborative effort that involves participants from several countries (notably Japan, where manga is a popular art form, and a vibrant market).

Conversely, a whole portion of the conference was dedicated to discussing inclusive access, head-on. Notably, Cristina Mussinelli (LIA Foundation) provided an overview of the Marrakesh Treaty and the European Accessibility Act. Avneesh Singh (DAISY Consortium) talked about the global accessibility specifications that are used to implement the legislative requirements.

To conclude, here is a quote from one of Avneesh Singh’s slides:

We are in the beginning of a revolution, we have a lifetime opportunity to achieve the era of born-accessible publications.
Let’s come along for a common worldwide standard for accessible publications

We certainly are. We certainly should.

Avneesh Singh talks about the "born accessible" publishing revolution.

This article was kindly submitted by software engineer, Daniel Weck, who works with the DAISY Consortium and EDRLab to implement accessible open source production tools and reading systems. Daniel also contributes to the standardization of web technologies and open publication formats, with a particular focus on inclusive publishing.

Notes from Berlin: DPUB Summit Highlights

Ken Jones, author of this article, displays his conference pass, program and conference information
 This is a guest post by Ken Jones – Ken specialises in writing workflow applications and offering training and consultancy for publishers on print and digital workflows. Ken’s company Circular Software’ provides software tools and services for a range of illustrated book publishers. Contact Ken on twitter @CircularKen on LinkedIn at linkedin.com/kenjones  or through the website circularsoftware.com.

This article has been kindly edited by Laura Brady and has been cross-posted on epubsecrets.com


The Digital Publishing Summit was held in Berlin on Weds 16th and Thurs 17th May. It was a wide-ranging and stimulating event full of interesting news about the digital publishing landscape.

The area of central Berlin we were in had a pleasant bohemian vibe. After my vegan panini and smoothie, I strolled to the event hotel as hordes of cyclists whizzed by along the sunny streets. Oh, wait, this is not a holiday review, allow me to continue… #not-eprdctn

This previous name for this yearly gathering used to be the ‘EPUB summit’ but this year it has been renamed the Digital Publishing Summit or ‘DPUB summit’ as it relates to digital publishing beyond EPUB. Before the main conference started around 40 of us started off the Wednesday morning with an extra Readium workshop led by Laurent Le Meur, CTO, EDRLab (@lmrlaurent). Laurent started but outlining some of the challenges with Readium 1 — notably performance, accessibility gaps and lack of Windows and Linux SDKs. They decided to  to start R2 with fresh codebase and to create new mobile & desktop SDK named Readium-2 (aka R2).

The overall aim of the project is to deploy a set of open and interoperable digital publishing technologies in Europe, around an open, flexible & accessible standard applicable to all kinds of digital publications

Laurent Le Meur presenting at DPUB Summit

R2 is being optimised for performance, ease of maintenance, and is well documented with consistency between target environments (so different OS, mobile, desktop). It is a live ongoing project with releases approx. every two weeks. Everything they are doing is on Github. Readium 2 apps are emerging already and an Alpha desktop version in publicly available.

They have an ambitious roadmap for 2018. Expect progress to be made in roughly in the following order: import, library, reflowable EPUB, LCP (Readium Licensed Content Protection), user settings, RTL, fixed-layout, CBZ (comic books and manga), OPDS (Open Publication Distribution System), bookmarks, vertical text, themes, footnotes, accessibility navigation, search, support for WP, audiobooks, TTS, media overlays, more dyslexia support.

Laurent Le Meur presenting the EDRLab Roadmap for 2018

 

EDRLab are seeking a wider community for development, feedback, and financial support.

Phew! Got all that? I’ll explain in more detail about some of this below. The roadmap is available on GithubEDRLab are seeking a wider community for development, feedback and financial support. They need constructive testing (so don’t just moan about how it may look wonky or has missing features, they already know this!…) and, again, everything is open and logged on Github.

Hadrien Gardeur of Feedbooks then spoke about the Readium Web Publication Manifest. Think of it as an unpackaged and deconstructed ebook that exists in RAM. By abstracting the contents it means R2 reading apps can be much more responsive. By loading the publication into memory and preparing for transmission over the web. And it doesn’t have to be EPUB, any structured publication could be deconstructed. They already support CBZ (a format for comic books as images) and are in a good place to handle any new W3C formats that come in the future.

R2 ingests packaged EPUBs in this way ready to them streamed, parsed, and fetched as HTML resources. They refer to the flow of rendered content to the reader app as the ‘streamer’. Think of the R2 reading system in 2 parts: frontend and backend. The frontend reader is fed by the backend, which could be either within an offline app or streamed from a server. By pre-fetching and pre-rendering content in backend memory everything is presented super fast.

Readium CSS is the open resource that Readium apps are using to present content and to handle user settings. Glue JS is a new project being built for R2 which makes sense of pagination, scrolling, touch and key events and CSS, locations, and custom properties.

Aferdita Muriqi of EDRLab (@AferditaMuriqi) talked about where they are with Readium mobile apps and their future plans.

Aferdita Muriqi of EDRLab (@AferditaMuriqi) talked about where they are with Readium mobile apps and their future plans. The R2 iOS app is out now and it is still being developed with later builds regularly and betas here. The R2 Android app is also available now (in beta) from Googleand Github. R2 iOS and Android are being developed in tandem and share the same functionality and core code. But the different OS mean they have to handle things differently.

It’s important to understand that the EDRLab stance on mobile apps is that they will only ever be ‘test apps.’ They have no plans to produce a fully featured mobile reader instead will provide an open source example to others of what is possible and exactly how to do it.

The ultimate goal is a robust attractive reading system and to become an open replacement.

Daniel Weck of EDRLab (@DanielWeck) talked about Readium desktop apps and their future plans. Unlike the mobile apps the plan for desktop IS NOT to be just a test app. The ultimate goal is a robust attractive reading system and to become an open replacement. R2 desktop apps are already very impressive but are still in early stages. Latest builds available here.

Video demo of Readium desktop app

R2 apps (desktop and mobile) already come with support for OPDS (Open Publication Distribution System) basically a way for a metadata feed from a vast library of books to stream and also can point to book content. Sideloading is also possible. R2 Desktop has a load button for local files and mobile apps can load from weblinks or Dropbox etc. via the OS. If you interested in the future of EPUB readers, or the current lack of them, consider join me and others in testing and giving feedback to shape these valuable and open source mobile and desktop apps.

If you interested in the future of EPUB readers, or the current lack of them, consider join me and others in testing and giving feedback to shape these valuable and open source mobile and desktop apps

We then moved on to an interesting chat about EPUB Canonical Fragment Identifiers. In short, EPUB CFI is an accurate way to point to any content within EPUBs. Lars Wallin of @ColibrioReader demonstrated how to use EPUB CFI to achieve interoperable, sharable, annotations, bookmarks etc. EPUB CFI also support identification of interactive elements, images and even parts of images within EPUB. Yuri Khramov of @EvidentPoint then spoke how they are contributing to ‘Readium NG’ to support existing R1 users in their migration to R2 with as little pain as possible

Later, in the main DPUB conference, we heard more on how Readium Licensed Content Protection aka Readium LCP can store a user’s passphrase inside an EPUB reader to effortlessly unlock content. Hermann Eckel of German ebook company Tolino explained how Readium LCP is already installed in their hardware and how it is now their preferred copy protected solution.

Later we saw a live demo of a reader entering accessing an ebook using the standard library log-in.
Slide displaying details of the readium 2 LCP project

Eden Livres talked about how they invested last year to integrate LCP server side and now have over 80,000 titles now live with Readium LCP.

De Marque talked about how National Libraries Quebec achieve 6M+ ebook loans – quite a lot for a small city. Previous DRM systems had cost them over $800K CAD over the years and with countless hours of support and left them feeling powerless and not in control. LCP was a community project to take that control back. They are currently focus testing LCP with user in Quebec City, to be released this summer for all public libraries in Quebec province and other partners.

“DRM should be boring. It should be invisible and nothing to the user.”

A refreshingly neat quote that summed it up was “DRM should be boring. It should be invisible and nothing to the user.” #eprdctn

Garth Conboy of Google spoke about EPUB3.2 and future plans for EPUB at W3C and invited participation in the community group.

Slide displaying details of how the edge browser supports EPUB

Ben Walters of Microsoft (@_BenWalters_) gave a great presentation on how their Edge browser natively supports EPUB. Double click any DRM free reflowable or Fixed Layout EPUB on Windows 10 and it auto opens in this well featured reader. He said that reading books on a desktop or laptop may not be common but actually for reference or education the ability to have multiple pages of the same book, of different books, or other web content side by side can be very useful. One challenge for browser based reading is that people don’t expect a browser to work offline, even though they can for cached and local content. The challenge here is in education.

Microsoft Edge for iOS and Android edge on phone coming ‘very soon’. First releases were last year but last month’s update brings improvements including notes, better navigation, page sharing and media overlays! This is great stuff from Microsoft!… You hear that @Adobe…

Ben confirmed to me afterwards that, at launch, sideloading will not be possible for Edge on Android and iOS. It will only feature content from the Microsoft store.

Ben also won the prize for the most amusing slide when introducing his agenda.

Amusing agenda slide from Ben Dugas

Stephan Knecht from Bones AG demoed how their audio reader accepts EPUB by parsing a book into txt files. Able to handle cues like SSML, to indicate preferred speech speed, male, female voices for example. More here.

Stephan made an interesting point that his customers (often visually impaired, elderly, veterans) like the static simplicity of dedicated hardware. Rather than making a phone app, his credit card sized audio player has a run time 40 hrs with standby of months.

Word files are a ‘post paper digital artifact’

I really enjoyed the talk from Adam Hyde of Coko Foundation. He explained in a friendly rant how 90% of scholarly content is supplied as Word files and not truly digital. His phrase was Word files are a ‘post paper digital artifact.’ But not just there to complain, Adam had three solutions that all looked worth finding out more about:

  1. Getting content from word into HTML – XSweet.coko.foundation was released two weeks ago.
  2. How to edit content in HTML including citations, notes, tracked changes etc. (that is, the pro tools that are missing from Google Docs) — An open source HTML word processor.
  3. How to get digital content it out into legacy formats like PDF —Automation and using the browser as a typesetting engine: pagedmedia.org

Other highlights included @gleephapp a phone app that features a real-time barcode scanner of ISBNs and even book spines to make the link from physical to digital and then allows recommendations. This app has just launched in France with 33,500 users and 20,0000 titles.

Volker Oppmann of mojoreads (@onkelvolker) showed their neat book sharing platform which actually pays its users 10% commission for any sale that comes from their recommendation. #eprdctn

We also heard that Readium has further goals including an update to the Readium CloudReader with support for audiobooks.

I’m very glad I made the trip and I hope to attend again next year but I must conclude by saying this: it amazes me that Adobe is not attending, supporting, contributing development effort and funding to EDRLab. Privately they would have so much to gain themselves from this modern open source reading system And publicly, surely Adobe putting a couple of full time developers in EDRLab Paris for a year or two would both demonstrate and garner respect and support of their paying customers and the wider publishing community.

I’ll be talking more about Readium and demoing R2 during my PePcon session in a couple of weeks time if you’re attending CreativeProWeek and into EPUB and ebook development.

Also feel free to contact me directly if you’d like more info and I’ll help where I can.


Postscript. Typescript (@typescriptlang) seems to be a the current programming language of choice. Microsoft, EDRLab, and many of the smaller developers presenting at DPUB were all using it and speaking highly of it. http://www.typescriptlang.org