I've been taking a look at WebDAV recently as a means for enhancing some of our content management features.

WebDAV is an extension to HTTP that provides facilities for distributed authoring and versioning of documents. The protocol is natively supported on both the Windows and Mac desktops thereby allowing direct access, publishing and authoring of content held in remote repositories in a seamless way. To the end user the repository looks exactly like a network drive and they can use normal file management options to manage documents.

After reading a blog posting from Jon Udell yesterday and in particular the notion of WebDAV proxies and virtual documents, I got to wondering about where else WebDAV could be applied.

Central to WebDAV is the a notion of a hierarchy of documents, each of which can have its own metadata. Sounds very much like the typical browse path for academic journals to me: titles, issues, articles.

So what if we were to expose journals as a "virtual file system" using WebDAV? A user could then integrate a journal (or collection of journals) directly into their file-system. Accessing content would be as simple as browsing through directories (i.e. an issue) to find the relevant content (e.g. a PDF file, HTML file, etc).

Obviously access control is an issue here. Arbitrary users wouldn't be edit documents, so the file system would be read-only. And, apart from Open Access titles, not everyone would necessarily be authorized to access the full-text of all content within a journal. But that's OK too; a WebDAV server doesn't have to expose a real file-system. What's exposed can be a virtual collection of content and that content can be limited to just those journals, issues, and articles to which the user has access. In a similar fashion the WebDAV proxy could also ensure that usage statistics are accurately recorded.

There's other options too, e.g. for publishing and sharing related research data, supplementary information, etc. Authorized users (the publisher, author, editors) could add the related information directly into the WebDAV file-system.

It seems to me that this could be a pretty powerful technique. It would provide a simple, familiar metaphor for accessing journal content. And users wouldn't need to accumulate and organize local copies of PDFs (and they do!). Instead they would be able to mount the content from any networked computer.

The are some obvious downsides. There's no search option, no sophisticated browsing, etc. Viewing journals as a virtual file system of content is certainly not right for all users or usage patterns. But one thing that Web 2.0 has taught us is that supporting a flexible range of access options is an important criteria.

The idea speaks directly to a wider debate about how academic publishers could or should disseminate content in the future. Assuming the correct access controls are in place and the content metadata is easily and widely available, do publishers really need to more than expose their content in these kinds of ways? Instead leaving end users and intermediaries to add value?