For the past few months, we’ve been laying the groundwork for a new nypl.org to match the new structure of the Library. NYPL’s website is at this point about thirteen years old; that’s thirteen years of changes, elaborations, tweaks and idiosyncracies, all sitting in flat HTML files (with a few ColdFusion apps powering things like the events calendar). With everything else going on, we realized last year that this might be the only opportunity for a long while to really shake the Etch-a-sketch, starting with a blank slate and building the site (and infrastructure) we really want for 2008.
One core requirement was that we shift to a content management system, rather than continuing to build HTML pages; one of the big goals of the Digital Experience Group is to enable staff across NYPL to engage more actively with the web, and continuing to have a web production group that was the sole gatekeeper and publisher of web pages simply doesn’t scale. In the era of blogs and Facebook, online users are increasingly used to publishing their own information, and NYPL librarians should be no exception.
After looking at a number of options, we settled on Drupal for the platform: while the learning curve is steep (at times painfully so), Drupal is incredibly flexible, and seemed to have all the functionality we might need. At the same time, we’d noticed that the burgeoning Drupal-in-libraries community was reaching a critical mass – sessions on Drupal at Code4lib or Computers in Libraries were jammed to overflowing, a new mailing list was set up, and there’s been increasing buzz among library geeks. So, early this year, the Labs group dug in. After a few months of the aforementioned learning curve (which still smacks us in the face on a regular basis), we started to understand how the system works, and it turns out to resonate with our mindset in a number of ways.
For example, while working on the Digital Gallery redesign, the point I hammered over and over was that we needed to design from the item-level pages up, rather than the home page down; we know from site stats that more and more visitors are coming in via search engines and external links, landing directly on internal collection or item pages, so much so that the vast majority of the user experience is there. Pretty home pages may seem shiny and important, but the real meat of a site is underneath.
We took the same approach to rethinking nypl.org; because Drupal encourages us to think in terms of Node Types, we started off by thinking about the categories of entities we’d need to represent on the site. A bit of brainstorming later, we’d settled on the following:
- Locations
- Events
- Collections
- Services
- People
Interestingly, these represent the basic questions a user might ask (where, when, what, how, and who, respectively), and work quite nicely as the site’s top-level navigation.
That’s not all, though; a huge part of the Library’s digital strategy is the leveraging of staff expertise (we spend so much time talking about digitizing our collections, but comparably little on how to get our staff’s knowledge out online), and we need some mechanisms for doing so. The first one’s been around for a bit: a handful of librarians are blogging, and we’re scaling that up in coming months. However, the ability to write discrete posts is just one of a number of genres of what could be called “librarian-generated content” that we’ll want to build in coming months:
- Posts
- Links (a la del.icio.us)
- Guides
- Curated Exhibitions
- Audio / Video
Because of the taxonomy that’s baked into the heart of Drupal, these bits of librarian expertise will be hung off of nodes from the first list. So, someone might write a blog post about their branch, tagging the post with that location, while someone else might create a research guide hooked to a public program. We’ll be able to slice and dice these different kinds of information in the interface, so that for example an event page might contain relationships to both a location and some items from the collection, as well as a series of blog posts and a few external links. Thanks to the magic of tagging, lists of posts become a subject blog, while similarly-tagged links become recommendation lists (think Best of the Web 2.0).
One last thing: in that top-level navigation, add in a sixth type of page, which we’re calling “Topics”: here, we’d take advantage of the *incredible* strength of library cataloguing to aggregate together all of the other kinds of nodes that’re tagged with a common subject with materials from the catalog, digital repository and elsewhere that are catalogued with the same subject. There’s a huge amount of interface design work to figure out how to present that much heterogeneous information in one screen, but that’s what we’re gearing up to tackle once we get the foundation laid.
So, that’s the big picture; our hope is to start pushing changes out to the public site sooner than later, but we’re hammering out concrete timeframes right now (more on that in a week or two). The implementation of Locations and Events in Drupal is mostly done, and we’re diving into the interface design now (Michelle should have some things to show very soon)…