On Wed, Apr 27, 2011 at 7:25 PM, Naman Muley <naman.g.mu...@gmail.com> wrote: > > > Correct. But again, google does it globally. I am asking for a local map. > Because Google can't get into my sub-routers. Most of the campuses > now-a-days are hidden behing NATs. Google shall be able to map same location > for all people inside that network. My own insititute is a classic example. > Moreover, the point still remains, that if I am in my university campus or > my Company campus, or to put it generally, in a LAN, I am more interested in > finding locations of people on my LAN, inside the LANÂ only and not globally > like Google Latitude. For e.g. a very practical example is, If I am in the > library and I want to ask a doubt that I have regarding a concept to my > senior, I will need his relative location e.g, is he/she in the hostel, in > the library itself or not; coming to know that it is in the same institute > (Google latitude output) is of really no use. > The method that Vitor pointed out is the one I want to use. As I had > explained in the IBM example, if the security guard is 12 mts away from the > WAP at the main gate, I can obtain this information. Though I doubt Google > uses cellphone towers, I shoudn't think google has access to telephone > resources as they are contracted. The telephone base stations are contracted > by telephone service providers and do not relay free information. Vitor, the > hexagonal cells based method is used currently in the cellphone system. I > think for cell-phones, Google uses GPS and Wifi. If you really think, its > one of the research areas of wireless ad-hoc networks, to adopt this > hexagonal cells implementation in the ad-hoc computer network. > Overall, WAPs / Routers etc are the main components of LAN architecture. A > combination of these is generally used to create the wifi infrastructure of > campuses. Using Reverse DNS methods for wired, and location estimation using > signal strength from WAPs for wireless are the Ace methods through which I > am imagining the application to work. > I may be wrong in my summary, please correct me. > -Naman >>
The idea of being able to do more granular location is definitely interesting. What I would prefer more than an application though is some kind of service level module (Telepathy maybe?) that will do this. Use available access points and maybe something like server-less XMPP over Zeroconf to provide location details to others in the area. This way, an application like Marble could plot the location on the map, whereas an IM client could suggest 'foo is currently on the floor above you, you may want to talk face to face' and then seamlessly switch from facebook chat to local xmpp to you going and talking to him in person or whatever. I do not see much scope in making a separate application just to see where someone is especially when there is a plethora of other mapping applications. How much percentage of KDE users would want to launch an app like that? But providing a optional service module (which may not even have a distinct public interface) which somehow exposes that data to all applications allowing them to do something innovative with it would definitely be cool. Nikhil >> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<
