I'm a big fan of the work Nature publishing group is doing in experimenting with new methods of publishing, such as their iOS apps (which inspired me to "clone" their iPhone app) and the Encode app with the concept of "threads". But there's an aspect of the iPad app that puzzles me: Nature's app doesn't know that a linked article elsewhere in Nature is also available using the app. To illustrate, consider this piece on horse evolution (doi:10.1038/nature.2013.13261) which is available in the iPad app (with a subscription):
This News & Comment piece links to the main article (doi:10.1038/nature12323), which is also available in the app:
But if I click on the link in the app, I get the article displayed in a web browser view (just as if I was in a browser on the iPad, or any other device):
This is odd, to say the least. If the article is available in the app (and much more attractively formatted than on the web) why doesn't the app "know" this and take me to that view?
In a sense this is part of a larger problem with publisher-specific apps. If the article isn't part of the publisher's catalogue then you are off to the web (or, indeed, another publisher's app), which makes for jarring reading experience. Each web site (or app) has its own way of doing things). Part of this is because different publishers represent different silos, mostly locked behind paywalls.
We can extend this argument to links to other cited entities, such as sequences, specimens, chemicals, etc. In most cases these aren't linked, and if they are, we are whisked off to web site somewhere else (say GenBank), a web site furthermore that typically knows nothing about the article we came from (e.g., doesn't know about the citation relationship we've just traversed). I think we can do better than this, but it will need us to treat links as more than simply jump off points to the wider web. For example, if the Nature app not only knew about all the Nature articles that were available to it, but also stored information about DNA sequences, chemical compounds, taxa, and other entities, then we could have a richer reading experience with potentially denser cross links (e.g., the app could display a genome and list all the Nature articles that cite that genome). Of course, this is still limited by publisher, and ultimately we want to break that silo as well. This is the attraction of starting with Open Access publications so that we can link articles across publishers, but still navigate through their content in a consistent way.