tecznotes

Last week, we posted a few images from our ongoing Mappr project:

Over the past few days, word has been getting around. We've been mentioned by a few people I respect tremendously, including Adam Greenfield, Andy Baio, and the Flickr folks themselves. The Mappr Group is humming along with over 100 members and a slew of great suggestions for improvements.

Notably:

• Awareness of a world outside the United States. I'm working on it, hoping to solve this one in a way that supports the future addition of arbitrary place designations like "Disney Land" or "The Eiffel Tower." Dealing with synonyms is hard. Dealing with inferences based on multiple tags is hard. Accounting for place names split up among tags (e.g. "San", "Jose") is also hard. Well... hard in the sense that I'm still trying to make this work without every nontrivial database query pegging the CPU for half a minute. :)
• An open API. This is on my list in a huge way. My goal is to model this closely on Flickr's API, so there's not a lot of duplication of effort, and users can use our API to augment theirs.
• User-defined maps. I bit a chunk out of this one tonight by solving a linear algebra problem that has been percolating for the past week, thanks to a well-placed hint from Cassidy.

I've also been plunged into an interesting learning experience in facets and ontologies, through helpful prodding from Boris Anthony.