Research papers should be offered in e-book format
The conference committee for UIST 2011 decided to once again offer the proceedings on USB flash drives instead of providing a printed proceedings. Because, y’know, print is dead. One minor oversight: people are starting to move on from USB drives as they move to tablets. We had multiple people ask for a way to access the proceedings from their iPads (since ACM had not yet posted the papers in the Digital Library.
While a simple solution to allowing tablet users to read the conference papers would be to make them available on a local download site (a solution adopted by MobileHCI 2011 this year), thinking about the shift from reading papers on laptops to reading them on tablets got me thinking. Conferences (and journals) overwhelming rely on PDF. While PDF is a standard format these days, it’s a static format designed primarily for printing. In an age of dynamic publishing on the web, preparing “camera ready” PDF files is starting to look like a bit of an anachronism.
We could instead allow authors to submit “papers” as HTML5/CSS3 sites, but that might be a little too bleeding edge. But what about allowing authors to submit their research “papers” in e-book formats? Both EPUB 3 and Amazon’s new Kindle 8 Format are incorporating HTML5 and CSS3 elements, which means that researchers could create publications that support rich content and interaction. And they’d work great on tablets. We could still support “classic” PDF formatting for legacy purposes, but EPUB and K8F would allow explorations into new ways to share research.
Of course, we’d need a guinea pig conference willing to test the waters. Surely there’s an existing conference on digital publishing that would be willing to explore new systems for academic digital publishing. Anyone have pointers?
jeff pierce 
Definitely should be done. But what does the LaTeX toolchain for producing these formats look like?
Apparently for LaTeX the flow is -> HTML -> EPUB.
Would love to see more experimentation with publishing formats / capabilities, but, for now, I think backward compatibility (e.g. for sticking in the ACM DL) is pretty essential. Publishers should re-think the whole end-to-end experience of research “papers” thought, in fact, I just spoke to some folks at Elsevier who are working on the “paper of the future” though they’re pretty focused on life sciences / medical journals.
Yeah, the current model is an anachronism. It’s desired around a world where publishing is hard and/or expensive, content is presented in pages, content is static, context is fixed once “published”, content is intended to be printed, etc. Now that those factors no longer hold, there’s an opportunity to rethink many of the assumptions we hold around publishing.
Of course, things are complicated because we’ve conflated publishing with “impact” (since it’s easier to count “published, peer reviewed papers” rather than actually determining impact. But that’s a post for another day…
You could go even a step further and just directly publish papers to the various e-book outlets.
Take a look at Smashwords, for example. Upload a .doc and you can sign up for distribution to many of the online platforms including Kindle, Pubit, iTunes (iBookstore), Kobo, and many others. It’s easy as long as the formatting is kept simple.
And it’s free.
Both individual papers and collections of papers could be e-published.
I think something like this is where we’re likely going in the future.
(FYI: Publishing direct to Kindle and Pubit is possible as well. For iTunes, you can only publish direct to it from a Mac. But Smashwords makes it fairly easy to hit them all.)
I agree that publishing e-books directly is an interesting path (you could imagine researchers sharing preliminary results on their websites and then providing more polished versions via short e-books (like Kindle Singles). But before that happens I think we need a few things to change, the hardest of which is that publication has become a mechanism for keeping score rather than a mechanism for disseminating research results. Academic researchers won’t shift to publishing papers directly to e-book outlets until we come up with a new way of measuring impact or contribution that doesn’t depend on getting content accepted by an established conference with a sufficiently low acceptance rate. Perhaps we should separate the scoring and publication mechanisms completely: assemble conferences that don’t publish, but instead take existing published content (web, e-book, whatever) and select a sufficiently small (for those who feel their conferences are polluted by accepting too many papers) subset for presentation at the conference.
We could also really use a better way to distribute updates to published books. Why don’t Kindle (and other e-)books continually refresh themselves as readers discover and fix (possibly through a suggest / accept mechanism vetted through a publisher) typos and other errors?