I just returned from TOC 2013. I got the chance to catch up with colleagues and friends, as well as meeting new ones (and since I work remotely, I even got to meet some of my Safari colleagues IRL for the first time!)
The programming for this year’s TOC offered a few high points, as well: the “Get Better at Git: Applying Version Control to Publishing” session, run by Matthew McCullough and Tim Berglund of Github, provided me with a long, long overdue a-ha moment for using Git; and as a digital comics geek, I was thrilled (if you’ll pardon the pun) to see the legendary Mark Waid deliver an engrossing demo of his fantastic Thrillbent comics platform.
One of the sessions that I found most compelling was Alistair Croll and Hugh McGuire‘s “Book as API” talk. Hugh has covered the gist of this talk over on the O’Reilly TOC blog, and the whole post bears reading and thinking on—it’s compelling stuff:
If we start to think of “books as data,” then the traditional publisher’s role starts to sound a lot like the role of providing an API: A publisher’s job is to manage how and when and under what circumstances people (readers) or other services (book stores, libraries, other?) access books (data).
During his talk, Hugh focused on the indexing of content from a book and making that information available via an API, and called out particularly clever and interesting uses for this information, from one-off projects like Dracula Dissected (in which Bram Stoker’s novel, Dracula, is broken down into parts — people, locations, journeys, journal entries, letters, etc. — that are presented to the reader over a Google Earth map, and connected with the story’s internal timeline), to full-on services such as Small Demons, which takes the people, places, and things mentioned in books and shows you their relationships to other people, places, and things. It’s fascinating stuff, and opens up the possibilities for how readers can engage with books.
All this talk of atomizing the book’s information into discrete chunks that could be rearranged depending on context got me thinking about streaming books, which is a concept that we here at Safari talk about a lot—in fact, Liza Daly delivered a presentation on this idea at the IDPF Digital Book 2012, and I riffed off of her work for a talk I gave at the Guadalajara Book Fair in November of last year.
A streaming book is a book that lives on a server in discrete parts, as raw assets, and is delivered to the reader over the network as a uniquely packaged collection of assets that respond directly to the individual reader’s particular usage conditions.
So for example: let’s say that we have a book that lives on a server, in parts: we’ve got our main text, translated into a handful of languages and semantically marked up, but otherwise unadorned; accompanying images, in various sizes and resolutions; styles and layouts for different contexts, such as mobile phones, low-resolution eink devices, high-resolution tablets, or digital broadsheets; supplemental files such as video or audio, also at various file sizes and resolutions.
Using mechanisms such as content negotiation, a device can send the server information about its conditions — “I’m a low-resolution eink device sipping low bandwidth in the mountains of Colombia,” or “I’m a high-resolution tablet in high-bandwidth Hong Kong” — and the server can then assemble and deliver a version of the book that is appropriate for the reader’s context: an image-less, plaintext version for our friend in Colombia, perhaps, and a high-res, finely laid out multimedia smorgasbord for our pal in Hong Kong.
Once you start thinking in this fashion, the possibilities become really, really compelling:
- A reader in Brazil can request a book on their browser, and the server can deliver a version in Portuguese instead of English.
- A reader on a mobile phone can get a version of the book which sports low-resolution images, and text that is specifically formatted for small screens.
- A reader can request the book in a version specifically designed for printing on demand, via either an Espresso Book Machine at a library or bookstore, or a copy shop service such as Paperight (one of the judge’s picks at this year’s TOC startup showcase).
- A reader on an iPad can receive a multimedia EPUB file, full of high-res images and widescreen videos.
- A reader on a Kindle can get the Mobi version of the book.
All this from one single repository (yep, still got Git on the brain), without having to create each version of a book manually each time — as long as the assets have been created correctly, are properly stored and described, and the server receives the information about a reader’s context, it can manage to serve up the correct version of a book to the reader automatically.
Moreover, using this approach, you can create books for mixed use within one space. For example, if a server knows that a request for a book is coming from a tablet, or a computer, or a TV, it can serve up different content for each context, thereby facilitating learning in a classroom setting:
the instructor gets a presentation-style layout for their wall-screen (the big board!); students on their tablets get a workbook-style layout with quizzes for evaluation; desktop computers get multimedia presentations and essay questions; mobile phones get shorter chunks of text, or surveys. All from the same source, and all on the fly.
Naturally, these techniques aren’t only appropriate for books — all types of editorial products can be thought of in this way. In fact, some already are: NPR treats its content in this way, and they enjoy a wide reach via various media as a result (for more info on this approach to content strategy, check out Content Strategy for Mobile by Karen McGrane, a short, fascinating, and incredibly useful read).
As ereading devices and services proliferate, it will become harder and harder for ebook makers to generate each necessary version of a book to reach all devices and contexts, and the process will become even more time-consuming and probably frustrating than it is now (I believe the technical term for this quixotic pursuit is “chasing the unicorn”). Approaches to content production and management such as the streaming book can help simplify the production process, and make it just a bit (or a helluva lot) more rational.