Compound tags make it easier to embed application-specific data into photos.
My friend Leo created a new group called One Word (http://www.flickr.com/groups/one-word/), which collects photos of single words. This is similar to the groups One Letter (http://www.flickr.com/groups/oneletter/) and One Digit (http://www.flickr.com/groups/onedigit/), which collect photos of individual letters and digits.
"Help!" Leo wrote to me, "What is a good way to tag the photos so they can be used for scripting purposes?"
Leo would like to enable people to make nice ransom notes [Hack #47] out of these photos, by identifying them with tags. Unfortunately, simply adding tags that match the words in the photos is not enough.
Why not? Consider the problem here. Let's say Leo adds the photo shown in Figure 2-13, of a gravestone containing the word PEACE, and tags the photo peace.
Figure 2-13. Leo's photo of a gravestone
There are lots of other photos on Flickr tagged with peace, and finding the one of the gravestone is going to be nigh on impossible unless you plan to write some fancy image-recognition software. And we really don't want to be wasting our computer cycles like that, do we?
"Aha!" you say, we can collect just the photos that are in the One Word group, using group membership as a filter. This will reduce the list to just photos that contain words. Or, we can add ...