The panel is led by Tim O'Reilly and features John Betz, Perry Evans, Kim Fennell, and John Hanke.
Tim is using the break time to talk about things necessary to build an Internet operating system. Location stuff. How should MapQuest, for example, get into this? It was a killer app but it's just one of many sites offering those services. It's not a winner take all thing like Amazon.com. Tim thinks that's because of there being no user generated content.
Keyhole.com demo going on. Very cool shit. Amazingly good high resolution data and a UI to control wandering through it. Traditional mapping data (and geo data) is overlaid on it. They have 10TB of data and are headed toward a 100TB model of the earth.
They've added user authored data to the mix too! Very cool. Users posts about location show up on the map. I must play with that later.
Tim has asked the presenters to introduce themselves and discuss what their businesses do, to illustrate the landscape.
Micorsoft MapPoint provides a web service.
Many web sites using IP based location services to lock their users out of content and services.
(Editorial note. This panel is not holding my attention very well. I'm not sure why.)
Tim thinks services need to figure out how to get user contributed data. MapQuest had TripNotes but AOL took it off.
Systems need more feedback loops now that we have location enabled devices like cell phone gps and triangulation. Need aggregation, tagging, and relevance for new services to work well.
Innovation in Google's and Yahoo's local services? Yes and no. There's a lot incremental stuff happening. Lots of local ad dollars. The next step beyond this is not just advertising--it's finding this stuff on the fly. That means the content providers and carries need to get together.
Cell phone triangulation. Is it being done as a web service and with standardized data? Microsoft with Bell Canada and Sprint to do some of this. GPS is part of the CDMA chipset, so there are millions of GPS devices out there.
Need permissions on location reporting too. Should apps be more aware of locations? How does that happen? Internet developers haven't really had to worry about the "where and when" of things.
Q: Cameras that location stamp photos?
A: Yes, but don't know model.
Q: Going beyond street addresses, like stores in a mall?
A: Yes, we need to go deeper into this.
Q: VOIP phones and GPS? Does that solve 911?
A: Sidetracked a bit. But not every device will have GPS. Tim thinks they will.
See Also: My Web 2.0 post archive for coverage of all the other sessions I attended.
Posted by jzawodn at October 06, 2004 04:05 PM