Felix Petersen on the geoweb (web2 expo berlin)

Super quick notes from Felix Petersen's talk last week at Web2.0 Expo in Berlin. Felix founded Plazes, a web-supported location-based service in 2006. He basically built upon his experience to describe the state of the geoweb in 2008 (excuse the rawness of my notes):

"founded in 2006 / one of the first start-up in this geopresence space check each others' location with your laptop that detects the wifi where you are we had to build the whole thing, now the geoweb is a more mature ecosystem and the components ar emore specialized nokia acquired plazes few months ago, now plazes is part of nokia maps; trying to mash things up

why going with nokia? if you look at my card it says "heads of social activities" my mother thinks i am organizing bbq party but it's a sign that nokia is advanced division " context-based services", expanded from LBA

the promises of geoweb: connected guidance of people (reach any place on any terrain... mass market... on foot+car), discover and share places (find and add cool places, recommend to others), record and share your life (record hikes, journeys and share or relive your life), stay close to your friends (stays close+ plan joint activities)

this promise has been around for sometimes the nokia phone from 1998 with WAP was cool and it even had a buddy-finder (vodagone germany) but it did not work, it never came out on that product broken promise, it was very naive, people seemed to haven't think lbs through

falses ideas: it's creepy that people can see where I am: i don't think so cool, i can finally track my friends wherever they are: this is wrong, it's not about tracking but about publishin great i can push an ad or coupon to anyone passing by my store: naive vision + this does not scale, this use case won't work

AND TIME WON'T HELP, it's not just about waiting the other part of the hype cycle

what it takes to build the geoweb, you need to understand that: - location detection: it's not soft, of course there is GPS enabled devices but... - mapping: layer of rules and metadata, google maps helps a lot, that's a component that is at a good stage - social objects: what are the geoweb specific objects that emerges: what are the unique objects that work with location?

1) location detection: GPS, A-GPS, wifi localisation (skyhook wireless), cell localisation, cell-triangulation (never really useful because it was confiné au seul operateur: pas possible de transmettre cette information aux amis d'un autre provider sans qu'ils aient a payer...)

what does location detection save? it can put you at a certain lat/long at a certain time does not sound like much, lat/long without context is not much but if you add existing social objects, location can add valuable information: geotagging pictures, twibble other objects: email, music tracks? what other objects can improve through location? sorting/filtering your email by location? this is where things will happen

there are still challenged: battery life: and it needs to be aware all the time... probing frequency privacy scheme needed pattern detection background processes: it's not possible to build an app that track my position in the background for other applications

there is a tendency to build gps in every devices (camera, bike) but interestingly there are also tracker that does uniquely tracking: little sony gps box

2) Mapping provides context and rules lots of improvement in that field enables routing and navigation personal nav is already mass market today mapping will constantly improve: google cars that collect streetview imagery, 3d models...

3) Social objects a social network is not only a connection between people, social connections happens through social object examples of social objects: bookmarks, people, pictures... web2.0 is the web of social objects

a social object: represents atomical data (i can refer to it, it becomes a conversational piece: like a picture on flickr), has a perman-linkable/embedable, often provides a conversational canva

you can add location to other social objects and sort them so which new social objects emerge in the geoweb? 1) places (a point in space): the most important, places are the most reference points of the geoweb the challenge: name space unification (how to refer to you couch as a place? your appartment? this is not just using the yellow pages, the idea is to come up with sth that is open but has a controlled vocabulary), open versus closes, dupes (we have to allow duplicates (different city names berlin) but also control them: how do you allow for ambiguity), lat/long which allows to place assignement yahoo had done it with the "where on earth" idea 2) traces: lots of points over time ideal container for attaching places and media 3) activities vast amount of data social activies are on the intersection of time and space what dopplr calls a trip is an activity

SO what can we build? 1) cool new apps: lat/long: this picture was taken here +place: three of your friends are at this bar +trace: this place I visited on my italy trip +activity: five people you will run into at Shift08

2) privacy right now: time and relation future: place-based privacy other options: chaffing, granularity

3) advertising - monetize targeting based on meaningful locations triggering upon engagement: you only get an ad if you engage physically a place, if you go there... turning advertising into information

what the ecosystem will look like? object creation (phone, laptop... automatic)...api layer... object storage (contextual logic/provacy, data layer)... api layer... object publication (published through all kinds of chanels: twitter, facebook, ovi, google earth, skype, fireagle...)"

Why do I blog this? material for my book about locative media, content for a foresight study about the future of the web.