semapedia.org is making the rounds recently. I was just fooling around with it on Friday, after having heard the folks on #mobitopia talking about it. There’s a post at the Social Software weblog asking if there isn’t a less obtrusive way to do the same thing. I was thinking about that as well. There are a couple of different aspects to make sure you have covered though. Here’s my brief rundown.
Systems somewhat like Semapedia.org, and by extension Semacode, have been around for a while. If we take the base idea as attaching some bit of information to a place, there are projects using Bluetooth to find nearby devices (which can include static beacons) and using GPS and triangulation based on Wifi access points or cell phone towers. What really stuck out about this one for me was that the information was embedded right in the tag. For most location based services you have some central server in the sky that takes the information about your location and directs you to the right resource. It centralizes control and restricts what people can do with the system. However, the semacode embeds the information for the resource right in the tag. No central control, no intermediating server, no need to get my information “put into the system” for others to look up. I just put my info online and I can print out the tags myself. Completely federated tagging. That was one of the big plusses that caught my attention when I first read about it. There’s no infrastructure that needs to be ongoing in order to support it. People need to have a reader installed on their phone, but that can’t be “broken” by a central server being down or taken out of service. Once you bootstrap this effort it stays up.
Now, of course we can have a Bluetooth device out there sending out a full URL in whatever it sends. Lets look past the problem that it isn’t meant to send URLs (you can always do things like attaching them as the homepage for a contact or something, or have a dedicated app like semacode does, so it’s not really much of a “problem”). There’s both the power and cost problems. It costs me almost nothing to create a semacode, I just went online and printed out a tag. If I already have a printer that’s a trivial cost. Totally the kinds of cost you want to see when you’re encouraging grassroots efforts. However the cost to deploy a bluetooth device to do the same thing is much higher. Plus it needs to be powered, which is problematic in both wired and battery delivered versions. And a device, even a simple one that just transmits some info, has value. I’m sure they would be stolen almost as fast as they get put out, especially if there weren’t tamperproof. People would steal them to reuse them for their own tags. And creating a tamperproof system really drives up the cost. However a printed sheet with one particular mark on it isn’t likely to be stolen. Vandalized or destroyed, sure. But no one’s goig to steal it. And replacing it is very simple. This is the deployment problem. In this respect as well Semacode really does a great job of fitting the kind of pattern that a community effort wants to follow. Semacodes turn any printer into a location markup device, and there are lots of printers out there already. Bluetooth basestations would work, but really fit the pattern of an infrastructure deployment more than a community effort.
I really hate to say it, but there is one other technology that could work well here, RFID. However, it would have to be a version of the tags that are capable of storing a full URL directly, and I’m pretty sure that passive tags aren’t nearly at that storage capacity yet. They’re meant to hold UPC style info. That doesn’t keep you from using multiple tags and some kind of encoding to spread the data across multiple tags. Hmm.. wonder if anyone is doing that. And then there are “active tags”, which are effectively little radio transmitters with batteries. The passive tags are much more appealing because they’re a lot smaller, cheaper, and specifically because they don’t have batteries. I don’t like RFID by the way. I think it’s a totally overhyped and underdelivered area of work. However, I would be somewhat interested if someone came up with a way to get low cost passive RFID tags to hold a URL of the users choosing and a system for using them to do this kind of community tagging. We could just wait for smartdust , and paint or pixiedust the tag on. Tagging pixiedust, I hope del.icio.us doesn’t have a trademark on that.