Title *

SummaryThe first rev of this sensing tool; titled 'citybot'. A description of the goals.

Description *The goal of Angel is an ambient sensor over a city. Citybot is more modest - it is a pared down subset of the bigger vision. This is the first pass at what I hope will become a real time trading service - it matches buyers with sellers using twitter. The history of this is that I've wanted to build a lightweight social bartering service for quite a few years now, in fact my older project thingster was intended to be this - but it predated google maps and I got bogged down in trying to render maps. Platial also I hoped would provide some kind of brokerage service also - but even just collecting and showing users favorite places took all of our time and energy. Part of the problem I think with many of the social place sharing sites was their silo nature. I tried to start an aggregator entitled placedb in my waning days at Platial but I just couldn't quite articulate that vision - and fireeagle does it now anyway. With Citybot I'm able to get further because I have to do a lot less - and the coverage of twitter is so large that I don't have to deal with aggregation. I can see that by building on twitter I should be able to get closer to the idea of a real time service for signalling about wants or haves. The risk of course is that the folks at Twitter will just do it themselves and kill the ecosystem. The kinds of things I hope to see people use the service for includes things like donating free time, or telling people that you have extra apples in your garden, or that you're happy to help volunteer some time for something, or that you're bored and want friends to hang out with right now - that kind of stuff. It's all the kind of stuff that we normally think about every day but that isn't very easy to tell other folks. To be more specific I see people using hashtags like so: * @locative the name of the agent itself for now * #all subscribe to all requests * #have you have something * #want you want something * #nearby it should be nearby * #bike a thing that you want or have * L:pdx where you are if not your default location on twitter * #private privately message me back please only * #public publically message me please * #free the thing is free * #money it costs money * #trade it is a trade in services * #silent do not send tweets to me * #digest only sent me tweets once a day * #realtime send me responses as soon as possible * #now time sensitive * #later not time sensitive * #anytime etcetera Why not just use Craigslist? I am very impressed with Craigslist - in fact it helped me find my dog only last week - but I still feel a need for something a little bit more precise and a bit more "in your face" than craigslist. Craigslist doesn't collect location data very well - I've talked to the founder about this but it just isn't a very high priority. And it doesn't really have a model of its user base - it can't signal to the people who express a specific interest... you have to search it - and the burden is placed on you - which seems a bit heavy. The funny thing is that the service is a way of adding "groups" to twitter... Effectively citybot is one group. I may go the route of letting people make their own channels; or groups; by spawning clones of the service or maybe just letting them create groups on twitter through citybot. Also, another funny thing is that it creates a social network not of explicit first order connections but one of geography and shared interest. Twitter really fails at surfacing second order networks - friends of friends. Anyway this is all the vision. Currently the system is very simple and very early. The first pass is just to surface the conversation - to show as much data as possible. To see if the idea is sound. The next pass will be to try to do some real analytics.