Last week Mark Wallace at 3PointD blogged Sloog, a new place tagging system native to Second Life developed by the sleek-looking Mosi-Mosi company in Barcelona. Their approach made me very happy, and I plied them with feedback. A tagging system like this is a far superior way of maintaining an inventory of interesting places in Second Life than, well, my avatar’s Landmarks folder in the inventory.
Mosi-Mosi has now added support for RSS, which allows me to highlight my most recent locations in the sidebar of this blog. I’d love to also be able to alter the title and add a description, but I think some of these features are on their way.
What’s interesting is that the original del.icio.us method, so effective for tagging and sharing places on the web, has also been applied to places in the real world. Tagzania lets you tag specific real-world locations via Google Maps and then provides feeds of your tagged locations, both as RSS and as KML (the file format native to Google Earth).
So now we have three tagging systems, one each for the three separate coordinate systems that we use to navigate — in the real world (latitude and longitude), the web (URLs), and Second Life (x/y/z coordinates, parcel name and region name). What’s next?
I’m hoping for mashups that connect all three of these spaces — meta-mashups, if you will. Take the House of Sweden: It has a physical location in Washington DC; it has a website associated with it; and the Swedish Institute will soon have a simulacrum of it at a specific spot in Second Life. These three places are obviously linked, but there is as yet no way of making that connection visible. Or another example: Harvard Law’s Austin Hall in the real world, on the web, and in Second Life.
In database-speak, it would be great to join the lookup tables of these three different tagging databases so that if you find one location, you automatically come across the others.
How to do this? Getting it mostly right may be possible just by comparing supplied tags in all three coordinate systems and joining the most similar ones. Otherwise, it’d have to be a collaborative effort, letting organizations themselves supply the requisite metadata.