Damon, Jon and I will present Project PlaceSite and discuss wi-fi cafe fun in San Francisco next week at Where 2.0, O'Reilly's new conference about location-aware tech.
If you're there, come say hi.
A bizarre media storm has gathered around wireless Internet cafes. Project PlaceSite and I have benefited. But this all deserves a closer look.
Tomorrow's New York Times quotes me in an article by Glenn Fleishman. My words appeared in a Seattle Post-Intelligencer piece last week. On May 30 a Financial Times article about wi-fi in cafes mentioned "zombie effect" [definition here], a term we invented to explain some of the reasoning behind PlaceSite. All this mainstream coverage followed Web buzz about an entry by Glenn on his Wi-Fi Networking News weblog. The entry announced that a Seattle cafe had tried turning off wi-fi on the weekends.
I'm thankful for the PlaceSite publicity but for the record: each of my partners, Damon McCormick and Jon Snydal, contributed to this project at least as much as I did. Professor Marti Hearst served a critical role as our project advisor.
A problem with the coverage: The Financial Times article strongly implies a trend in cafes across the country that involves reduction or removal of wi-fi access. But the opposite is true, at least in Seattle and San Francisco: wi-fi is becoming more ubiquitous in cafes. The article cites just three cafes -- one in Seattle and two in San Francisco -- that have limited their wi-fi access. But hundreds of cafes in these cities offer wi-fi service, and more cafes add wi-fi every month.
I see no evidence of a new trend: both of the San Francisco cafes in question have been experimenting with limited access for more than a year.
The other articles, particularly the New York Times piece, were more balanced and better informed about this. But I sense a media snowball effect that might trigger an avalanche of inaccurate coverage.
A warning to reporters: consider the numbers here, so you don't mistake aberrant behavior for what's clearly the norm.
[ UPDATE: Our launch period at A'Cuppa Tea is over. Keep an eye on placesite.com for news of upcoming launches. ]
We'll launch Project PlaceSite this Tuesday
at A'Cuppa Tea cafe and teahouse in Berkeley.
Full details: PlaceSite.com.
Come out and join in!
Imagine opening your laptop computer in a neighborhood wireless Internet café and firing up a Web browser. Instead of your usual startup page, imagine this on your screen:
That's the core of Project PlaceSite. It introduces a new way of using wireless networks -- to create a local information service by, for and about people who are in the same café together.
We're rolling it out in Berkeley in a few weeks. Details: www.placesite.com.
Please let us know what you think. And come out and take part!
Lars Holmquist spoke of "PacMan Must Die” at Intel's Berkeley research lablet Friday.
This is an innovative game developed by Holmquist's students at the Viktoria Institute's Future Applications Lab in Göteborg, Sweden. It's a tweaked-out multi-player version of the classic game Pac Man, with two major twists.
The first twist: characters’ roles are switched. Players control ghosts invading Pac Man's home turf, trying to recover the dots stolen by Pac Man in the original game.
The second twist: the playing field is distributed across two or more devices held by multiple players.
To finish a level, a player must eat dots not just on her own screen, but on the other players' screens as well. If you send your ghost through a doorway on the bottom of your screen, the ghost disappears from your device. It enters another player's screen through a corresponding doorway. The game allows up to five players to join in on the distributed fun.
Players have to look over at one another's screens to see where to guide their characters. Physical strategy and cooperation become central to this virtual game. Opportunities for new sorts of pranks arise -- for instance, you can physically run off with your friend’s ghost.
I love this; it's another way of combining video game fun with the fun of play in real-world places.
This is the sort of rich, simple innovation that I hoped would emerge with the wi-fi enabled Nintendo DS portable video game system. But Nintendo seems to have locked down DS development, limiting it to internal and professional developers. Such professionals have years of experience and training in building traditional games. This background cripples their ability to innovate, to see beyond the constraints of traditional game platforms.
Nintendo, learn from eBay and Google and Amazon: let customers and outsiders build value for you. Open your platform and let it thrive.
Why not use wi-fi (wireless Internet) access points to track down stolen cars, bikes, purses and other valuables?
Many of Earth's major cities are becoming saturated with wi-fi access points. It's hard to find a public place in San Francisco, for instance, where a wi-fi device can't detect a nearby access point.
Imagine placing a narrow wi-fi beacon device inside the frame of your bicycle. You tell the beacon that, every day at 4 a.m., the bike is locked up at your house. Next time the clock strikes 4 a.m., the beacon turns itself on and it makes note of which wi-fi access points it can "see" from your home. It remembers that these access points represent home. Then it turns itself off again. (Wi-fi detection drains a lot of battery power -- the device stays off most of the time to save juice).
Two days later, at precisely 4 a.m., the beacon powers on and notes what access points it can "see." If it detects one or more of the "home" access points, it turns itself off again. Two days later it does the same thing, and so on. We'll call this state of affairs the beacon's default mode.
During one of these early-morning access-point checks, if the device doesn't detect a home access point, it switches into "stolen mode." It powers on every 15 minutes and checks for any open access points. (Open access points are not encrypted, so anyone -- and in our case, any beacon -- can use them to connect to the Internet.)
When the device finds an open access point within range, it sends a message to a server. The message specifies which beacon sent it, as well as a unique identification number that specifies which wi-fi access point is being used. The message also includes a list of other access points that the device "saw" during its recent travels.
Software on the server connects to a database that stores the geographic locations of known access points. It uses this information to convert the list of access points recently "seen" by the beacon to a path on a map illustrating your stolen bike's recent movement. It converts the beacon's current access point to the beacon's current location, which it marks on the map and converts to a street address. The server sends this information to your e-mail account -- and to the police, if that's what you want.
(Each time the beacon connects to the server, it also checks for commands that it's supposed to follow. Via a password-protected Web interface, you can tell the beacon to switch from "stolen" mode back to default mode. You can set a new home location, and you can set a new time for it to check each day for its "home" access point fingerprint.)
This would provide essentially the same service offered by LoJack, an extremely expensive anti-theft system for cars. LoJack depends upon a beacon that's hidden in a car. If your LoJack-equipped car is stolen, when you file a police report a radio signal is sent to the beacon that puts it into "stolen" mode, which causes it to repeatedly emit a signal over a special radio frequency. Police cars and aircraft equipped with special LoJack sensing computers can track this signal and follow it to the car. LoJack is quite expensive (Carsdirect.com sells the device for $695, for instance), and the real expense comes from the service: LoJack employees and police have to be trained to use the tracking equipment, and so on. (Other services based on Global Positioning Satellite and cellular phone systems are also quite expensive.)
Now that wi-fi is almost ubiquitous in many areas, we can create a system that allows people who live in those areas to track their stolen goods without LoJack and all of its overhead -- at a tiny fraction of the cost.
But remember that new stolen-item-tracking technologies can be put to darker uses too. Such developments mean that almost anyone might have the power to plant tracking devices on unsuspecting people and vehicles.
Look for unplanned potential within the Nintendo DS, an upcoming handheld videogame system. It uses wi-fi, as well as a proprietary Nintendo wireless format, to let a player compete and communicate with another in the same room, or with a potentially infinite number of other players via connections to the Internet through wi-fi hotspots. The New York Times says the unit will go on sale in the U.S. November 21 for $149.
Nintendo's marketing, the press, and the weblogs all seem fixated on the fact that this unit has two screens, and the fact that it will let people play the same old types of multiplayer games in mobile settings.
But I think a special combination of attributes make this a potential source of compelling new smartmobbish applications and behavior:
(1) Not only does it provide wireless networking capabilities, it's built around using wi-fi in an ad-hoc person-to-person manner -- regardless of whether wi-fi hotspots are nearby.
(2) When an open wi-fi hotspot is nearby, it can be used to connect from the field with servers on the Internet.
(3) Kids will use it, and kids aren't locked into narrow wireless communication paradigms.
(4) It's not too expensive, which might encourage a critical mass of these things in urban areas. (Compare it with Sony's upcoming $300+ wi-fi-enabled PSP game device.)
Imagine the sociolocative fun that this might enable -- if Nintendo doesn't block out nonlicensed developers.
(Thanks to Matthew Rothenberg for the tip!)
Czechy doubt: Encounter Bubbles.
Imagine sharing the collection of MP3 music files that you listen to on your PDA wirelessly with anyone nearby. Imagine converting that PDA into your own mobile, wireless Web server, through which anyone nearby who has a wi-fi enabled laptop or device can browse and download whatever Web pages, photos or other content that you choose to offer up.
A new application called Pocket Rendezvous allows you to do that. This is exciting because it takes the mobile personal Web server paradigm (as seen in Intel’s Personal Servers and in Julian Bleecker’s Wi-fi Bedouin project), and rolls it out in a form that will run on mobile devices that thousands of people already use.
But get this: while Pocket Rendezvous uses the device discovery and networking protocol most famous for its use in Apple’s Rendezvous system, Pocket Rendezvous runs only on (Microsoft) PocketPC devices! How darkly ironic… Nonetheless, bravo to Simeda, the small German software firm behind Pocket Rendezvous. I hope they port this to PalmOS soon so I can use it on my wi-fi enabled Palm.
(Thanks to Joe, Howard and The Register for the tip.)
Julian Bleecker's wifi.Bedouin project has my mind churning. Bleecker frames this as a product and a service: essentially it's a laptop in a backpack with wi-fi antennas, a PDA remote control, and software that creates your own little "island Internet."
Forget about the packaging. The big innovation here lies in the paradigm, in viewing your wi-fi-enabled laptop as a server and a filter rather than a client.
What can you do with this? Here's an example: have fun in Starbucks. Walk into a Starbucks cafe, sit back and watch customers come in, fire up their laptops and connect to your wi-fi node. They think they've jacked in to the Internet, but really they're connected to your mobile server. You can serve their Web browsers whatever content you want -- an art piece, brand-damaging fake Starbucks ads, fake coupons, photos of your cat, whatever. Mix your content with real Internet connectivity and content served up via the cafe's wi-fi service. (Combine this with a Guerilla Cafe DJ setup and you've got a toolkit that would make Starbucks interventionist Reverend Billy proud.)
It's important that we engage in this sort of play and think through these things, because not all the possibilities brought to light here are funny. McDonald's or Starbuck's or anyone else can intercept passwords and can easily monitor, record, forge and censor unprotected wi-fi communications. We can prevent such misdeeds through technical means, but before the solutions can be perfected and adopted we need to raise public awareness that the problems exist. Pranksters can spread this sort of consciousness.
This is just one example of what we can do with systems like Bedouin. Check out Bleecker's scenarios page (and click through the three scenarios) for more.
3/24/04 UPDATE: Arthur Law brings up two other fun possibilities. (1) For business people and software developers: why not put the project work on a bedouin server and huddle the workgroup around a campfire? (2) For video game afficionados: won't weddings and funerals be more fun when you and your laptop-toting friends engage in action-packed shoot-em-up tournaments during the ceremonies? Why wait for high-speed Internet coverage to reach your destination when you can bring the connectivity with you?
5/04/04 UPDATE: I recently came across another intriguing application that converts local machines (in this case, handheld computers) into miniature wi-fi Web servers. It's called Hocman and it's designed to allow motorcyclists to exchange social information via HTTP when they encounter one another on the road. Details here.)
5/16/04 UPDATE: It turns out that Intel Research has been doing its own work using the mobile server paradigm, using tiny Personal Servers.
Yikes! How did I miss Microsoft's release of the World-Wide Media eXchange? Microsoft calls it "a centralized index of digital photos, where photos are tagged by the geographic location where they were shot." Developers affiliated with the Locative Media Lab and the Place Lab initiative have been discussing and working on very similar ideas for years, hoping to build out open and broad foundations for such systems before these capabilities can be locked down by narrow megacorporate interests.
As Jo Walsh put it, "interfaces and standards from meshed hyperconglomerates like Nokia and Microsoft present us with a square pinhole through which to attempt to view a potential wild and vivid world."
It's time to get cracking. Thanks to Scott Lederer for the wake-up call.
"Do I really need a camera attached to my mobile phone? Honestly, isn't this just a gimmick?"
Lately I've fielded those questions many times over from friends and family, and even from other tech people.
Even the phonecam manufacturers don't seem to have a clue what people will really use these things for, judging from the foolish scenarios they portray in TV commercials. But that's typical; new technologies are never born fully-formed. Nobody knows how networked cameras will evolve, and nobody knows just how we'll grow to use them. But special properties of networked cameras have convinced me that these tools won't be abandoned any time soon. Some of these capabilities haven't emerged yet but I think they're all on the way.
Here are five important capabilities that seem unique to networked digital cameras:
1) A photographer can use such a camera to send all her photos to a single, central storage place as she takes them. This eliminates the handling of film, smart cards and other intermediary media. It means that cameras can be smaller and cheaper because they don't need massive amounts of storage space. It dramatically simplifies problems involving backups, sorting, and after-the-fact annotation. No more rooting through PCs, CDs, servers, drawers and albums to find that great family portrait from last Thanksgiving.
No single firm or agency can or should store and control everybody's photos. Nobody's photos should -physically- be stored in just one facility. The media should be backed up and mirrored at multiple sites in case fires, floods or whatnot destroy the data at one site. But as far as the user is concerned, the photos should "live" in one secure spot in cyberspace. You should have just one virtual "place" to search through when seeking your photos, so that you don't have to worry about inadvertently losing important photos, and so that you don't have to constantly copy collected photos from one device or place to another.
2) Networked cameras provide the ability to annotate photos as they're created, or soon afterwards. Such annotation can include written comments from the field to accompany particular photos, but I don't think people will bother with that in most cases. Other forms of annotation might be more widely used and more important, including:
3) Networked cameras allow the capability to publish and share photos online as they're taken.
4) Most users usually have their phonecams on hand, just as so many people nowadays usually have a mobile phone on hand. In many ways this is great news: you'll be able to document all of life's wonderful and serendipitous happenings; you'll hardly ever say, "if only I had a camera at hand..." In Japan, where phonecams are old news, teenagers often snap shots for the purpose of enhancing face-to-face conversation. Most such shots are briefly shared via the devices' small screens and then deleted, never to be sent or stored anywhere. Of course this is bad news too; when everyone carries around networked cameras we'll all lose some of the privacy that we now enjoy.
5) Networked cameras provide the ability to transmit photos to audiences without permission from the authorities. These devices rob government officials of their ability to detect and destroy photographic evidence at border crossings. This capability will be especially beneficial in war zones and in police states.
I want to be a guerilla cafe DJ.
Here's what I need: a small, portable FM transmitter powerful enough to override the signal in any cafe that's playing megacorporate radio. This would be especially useful in cafes tuned in to that unbearable Clear Channel snooze-jazz station that has infected so many San Francisco Wi-Fi cafes.
When I enter a cafe and power up my laptop, I'd like to plug this magic transmitter box into my laptop and use it to overpower the cafe radio's reception of that cheesy station, replacing it with whatever music is emerging at the moment from my laptop's sound card.
If you know where I can obtain such a device, please clue me in.
(If you know who created the power-tower graphic above, please let me know so I can give them credit. I pulled this image from the Web a long time ago and I can't remember where it came from.)
This week's O'Reilly Emerging Technology Conference in Santa Clara, California was great fun. I enjoyed the presentations and the ideas, but it was the behavior of the attendees that really fascinated me.
The conference wasn't all "there."
Much of it took place elsewhere, and everywhere -- in cyberspace. My attention was always torn between the physical conference and the virtual conference.
I've never seen so many networked gadgets in use simultaneously in one place. During any given session, much of the audience had their laptops open and online thanks to power outlets and wireless Internet service throughout the conference rooms, lounges and hallways. I was immersed in bandwidth; I was surrounded by a chorus of whirring laptops and clicking keys.
For me, this was a totally new sort of event -- but soon, experiences like this will become commonplace.
The typical scene: up front the speaker presents her talk, projecting a slide show or a demo onto the wall-sized screens. A glance around the darkened room reveals dozens of ghostly blue-white faces gazing into laptop screens.
Many of them are engaged in online chat rooms. ConFab, a Web-based chat tool, was built just for the conference. It allows a person to mouse-over a map of the conference rooms, to specify which physical-world room he's sitting in, to engage in text chats with other attendees, and to see how many people are logged into each conference room. He can even pay a virtual visit to another conference room to find out what people are chatting about over there.
(Network problems made staying connected to ConFab very difficult. But people conferred in more traditional Internet Relay Chat rooms too.)
In the chat rooms people crack jokes and trade opinions about what the speaker is saying, and they write brief summaries of what's going on for people who are tuned in to the conference from other parts of the planet.
People read other folks' comments. They examine the speaker's Web site. They tune in to chats going on simultaneously in the other conference sessions, judging whether to step out and join the session going on next door.
And they blog. I watched at least three people pull out digital cameras during presentations, take snapshots and upload the images to their blogs right there.
People collaborate to take notes on the presentations and discussions using wikis. Groups of people use Hydra, a collaborative editing tool that allows multiple users to elegantly write, edit and add to a single document simultaneously.
That pattern was repeated endlessly throughout the conference. Everyone's energies were divided between cyberspace and the physical world. This is a fascinating phenomenon, but when the novelty wears off will such connectedness make for better or worse conferences?
Did the average attendee go home with more or less knowledge, with more or fewer useful acquaintances, with more or less encouragement than they would have acquired without the digital networking? What do you think?
The conference left me more confused about these questions than ever. For one thing, I wasted a lot of my attention and energy dealing with a couple of basic technical problems that the organizers can easily iron out in time for next year's conference. But next year, won't my attention be devoted to a new set of problems to wrestle or configurations to fine-tune as more real-world subtleties slip by unnoticed?
I want to experiment more with this, and I know I won't have to wait long.
(A freakish footnote: I'm writing this entry on my laptop in a Berkeley WiFi cafe, days after the conference ended. Three other geeks bend over three other laptops by the window. They're talking about their experiences at the same conference, as they post entries to their own blogs about it. Should I laugh or cry?)
(Photos in this entry by Derrick Story of the O'Reilly Network.)
People have predicted very complex "augmented-reality" systems that might arise in the near future, when many folks will carry around location-aware devices. But how about a simple thumbs-up/thumbs-down rating system?
Here's how it might work: your device includes a green thumbs-up button and a red thumbs-down button, TiVo-remote style. As you move through the city, when you enter a favorite restaurant or club or cafe you click the "thumbs up" button. When you pass that restaurant where you got food poisoning or that stuffy overpriced bar, you click "thumbs down." And if you enter an especially wonderful place, you click "thumbs up" twice to give it two thumbs up.
The key: you don't have to interrupt your daily activities; just reach into your pocket and click one of two buttons whenever you think of it.
Each time you press the button, the device records your geographical location and the thumb rating. Soon you have a little database, a map that shows the spots around town that you love and the spots that you hate.
So what? So nothing, until people share their preference maps. Now you have a powerful concept.
Thanks to this network, you can share your preference map with anyone who wants to use it, and you can freely use other peoples' preference maps. You decide which of your friends have tastes closest to your own, and you subscribe to those friends' preference maps.
Software on your device notifies you when you're near a spot that friends have rated positively; if a dozen of your friends rated a place highly, the device specifies via sounds or spoken words that the spot got a lot of thumbs up from the people you trust. And another thing -- software maps this for you, visually overlaying the green and red thumb-clicks over a map of a city or a region or a building.
You can also form preference groups, just as you form e-mail discussion lists. Everyone who shares an interest adds their account to a particular list, and that list compiles all members' preference maps into a master map for that group. Then anyone in the group can subscribe to the group map and use it or turn it off as desired. (Of course, if you no longer trust a person or a group's tastes, you can filter out their thumb-clicks on your map by removing that person or group from your list).
For certain events you use time-sensitive preference maps with thumb-clicks that fade over time. This could be great at an art fair or an outdoor festival -- you form a preference group with a bunch of friends who will attend the same event, and as you all explore the place, you each tag the coolest things and the most worthless things that you see. You might glance at a map and notice a dozen bright green blips at bandstand 3, which suggests that something amazing is going down there right now. Those green blips by the coat-check, on the other hand, have faded, so you probably missed whatever happened there. So you head straight to the action at bandstand 3.
(I'm reading the book "The Orchid Thief," and I just finished a scene that takes place at an orchid convention. Word spreads through the convention center that "you've got to check out the orchid that smells just like grape Kool-Aid." Frustrated flower freaks are milling around, blindly trying to find that particular orchid among hundreds of flowers on display. The orchid freaks would immediately know just where to find the most talked-about flowers in the show if they used such a preference system. That's what sparked this idea.)
I've been envisioning freakish scenarios that are bound to arise soon, when location-aware devices embrace payment and auction systems.
Imagine driving in circles, seeking a parking spot on a busy block in a crowded city. Your device immediately finds four people who are preparing to evacuate parking spaces within a few blocks of you:
Bill the fruit loop has plenty of time to spare and he lives for this crap; he's prepared to stubbornly hog that parking spot for hours, until someone meets his demands. What happens when Bill and his money-grubbing cronies take over all the public toilets at an outdoor concert? Not even our jiffy-johns are safe from speculating sleazebags.
I'm exaggerating, but the point stands: we'll face scenarios where a minority of selfish jerks can use location-aware networked technologies to manufacture new hassles in public spaces. Left unchecked, these are just the sorts of hassles that can draw more lawyers and politicians into the mix, just the sorts of hassles that can eventually prompt lumbering government bureaucracies to enact broad, boneheaded legislation that can hamstring innovation.
We should think through such scenarios now, and devise strategies for discouraging such abuse long before legislation and litigation rear their ugly heads.
(Thanks to headmap.com for inspiring this line of thinking. I love headmap because its authors spend a lot of time just imagining how location-aware technologies will fit into everyday life, then they write up their most intriguing thoughts and scenarios.)
Who are the people in the neighborhood? Have a look.
The geourl database brings real-world geography to the nebulous world of Web sites. Add your precise latitude and longitude to the database, then look up your online neighbors. (Tip: Here's an easy way to find your coordinates: enter your postal address here).