Last month Aaron Straup from Flickr announced four new API methods for machine tags (and also happened to mention our project!). One of these methods in particular, flickr.machinetags.getValues, inches Flickr closer to the idea of 'Flickr as database' for metadata storage.
In the past, if you wanted to find all the unique values of a given field (artist, for example), you had to query Flickr to get all the photos with 'artist' as a field, then make a call for each photo to get tag values...and then use a language like PHP to extract unique values. Once your collection gets past a few images, this sequence of events quickly becomes impractical.
With flickr.machinetags.getValues, you can get all unique values for a given field with just one call. This paves the way for creating UI links for browsing different values for fields. It also lessens the need to store metadata in a MySQL database, like we currently do.
As others have expressed interest in creating collections like accessCeramics, we've begun development on a more generic tool for cataloging Flickr images. We're also working on a companion presentation tool for viewing these catalogued collections. Thanks to the new API methods, the presentation software should run quickly, and shouldn't rely (much) on an external database. We'll post more information about this project as it becomes available.
Tuesday, January 27, 2009
Subscribe to:
Posts (Atom)
