Today's New York Times writes about a drive I took a few weeks ago with the editors of the latest edition of the American Institute of Architects Guide to New York City. The Guide is the definitive reference to architecture in New York City, with a strong editorial voice and a great sense of history.
The article mentions that I am their "web guru" because I built the content management system they're using to manage the photos and entries for the Guide. It's a project that demonstrates the real meaning and value of "Web 2.0." You have two editors on opposite sides of the Atlantic (Norval is based in France while Fran lives in New York). You have graduate student contributors. You have thousands of buildings to write about, and tens of thousands of photos of those buildings. You need to organize the buildings, select which ones will be included and which will not, upload and organize the photos, choose the good ones, and then produce several thousand pages of text. Sounds like they'll need a big custom system with a strong digital asset management component, doesn't it?
Well, no. What they are using took about a week to build. It's basically a mashup of existing systems like Google Documents and Flickr, glued together with a shared taxonomy and some custom programming.
They're writing the text in a shared Google document. No work required from me at all for that, but they have revision tracking, chat capability, and everything else they need to edit a document across the ocean.
The job of managing tens of thousands of photos, in a way that lets students and others upload photos easily, and lets the editors review them and choose the good ones, is left to Flickr, which is much better than I or anyone else will ever be at managing and storing photos, not to speak of providing easy-to-use tools. A private photo pool, and accounts for everyone taking photos for the book, and we have a "digital asset management" system more powerful than anything we could have built.
Now for the complex part -- associating the photos, the text, and all the metadata for every entry (when was the building built? Who was the architect? What's the style? What neighborhood does it belong to?). This is being done in a Drupal-based site I created for them, which has custom content types like "building" and "borough" and so on. Thanks to the Flickr API, I can use the building metadata in the Drupal database to pull the right photos from the Flickr photo pool, and the end result is a site where they can search and organize and tag buildings and see the appropriate photos.
A very large content management system, solved with existing tools, open APIs, and a few days' worth of programming. This is the Harmonica philosophy in action: simple, portable, flexible. Anyone can play it. It doesn't cost much and it fits in your pocket. But it's powerful, and most importantly, it gets the job done.
- Ken's blog
- Login to post comments