Receive new posts as email.
This site operates as an independent editorial operation. Advertising, sponsorships, and other non-editorial materials represent the opinions and messages of their respective origins, and not of the site operator. Part of the FM Tech advertising network.
Entire site and all contents except otherwise noted © Copyright 2001-2010 by Glenn Fleishman. Some images ©2006 Jupiterimages Corporation. All rights reserved. Please contact us for reprint rights. Linking is, of course, free and encouraged.
Here's an open forum for comments on airport electrical outlets: It's obviously a growing matter of concern, given the hullaballoo that surrounded Alex Saunders's post on Trudeau Airport covering (some) electrical outlets in Montreal. A colleague emailed from an airport in Paris noting that the airport magazine said 2,000 new outlets were being installed for travelers' use in Paris's facilities. Cory Doctorow of BoingBoing was accosted about using an uncertified (??) electrical device in bloody Luton.
Please post comments below about your experiences with finding electrical outlets at airports--good, bad, and ugly.
The Detroit News reports that the efforts to install wireless network access across Oakland County includes Cisco: The enormous scope of providing a variety of public wireless data services across this area adjoining the county that contains Detroit obviously worried some in the community. MichTel Communications was taking on an estimated $100 million project, including raising the funds, as no taxpayer dollars are being spent on capital expense. Cisco's involvement as a technology provider seems to have calmed nerves.
The service will offer free 128 Kbps symmetrical access and charge for higher speeds. They will also sell advertising and "naming rights," which in this context is a bit confusing. It sounds more like sponsorship of a splash screen, but I'm sure that will become clearer. The county has 1.2 million households and 40,000 businesses that will be passed by the Wi-Fi network.
The Lily Pad partnership in Cincinnati is volunteer and donation based: The network brings together the City of Cincinnati, Time Warner Cable, and Project Lily Pad, part of Give Back Cincinnati, a non-profit. The idea is to find sponsors who contribute small amounts of money to fund particular hotspots for an extended period of time. Avoiding any taxpayer dollars avoids a political debate and keeps the future more certain as legislation wouldn't affect this private operation. Time Warner Cable sponsors a large waterfront hotzone. Twenty locations are online now with 55 more in the works!
Singapore adds its first public park Wi-Fi: One-North Park offers Wi-Fi via the state parks department and a state-connected industrial park developer. Three hectares are covered now, with a total of 16 ultimately to have service. The park spans a number of developments. A parks press release from a few days ago notes, "When fully completed, it will serve as a seamless link connecting the key developments such as Biopolis, Fusionpolis, educational institutions and MRT stations. The Park will be the main recreational and relaxation space for the work-live-play-learn community in one-north."
Cambridge, Mass., extends MIT Roofnet: As disclosed several days ago, the Roofnet mesh-routing project at MIT will extend citywide across Cambridge. The city will install antennas on as many city buildings as possible, this article states. MIT's VP of IS&T said, "It won't necessarily be the highest speed, but it's going to be everywhere, just like air, and that's pretty cool," he says."
Alec Saunders writes that the Trudeau airport in Montréal has locked down its juice: The electrical outlets are apparently capped to prevent usage. He found a phone booth designed for dial-up modem users that has an exposed plug. Hydro Québec raising its rates?
The VIA Rail system in Canada has formally launched Internet access on the Windsor-Quebec City route: The service has been in testing by VIA Rail for quite a while; I wrote about it in July 2004 in a New York Times article that discussed Internet access on commuter transportation, including ferries, buses, and trains. The service will be in all first-class coaches by the end of April and all classes by the end of 2005. As reported a few months ago, all 22 stations will also have Wi-Fi service.
It seems that the reason for the extended trial may have to do with the Internet downlink. While I suspected that Pointshot and VIA Rail were waiting for 3G cell data service to catch up with the train's route, this article states that ground cellualr from Bell Canada and Telus was unacceptable. The service will use satellite access that will reach 3 Mbps in urban and 1 Mbps in rural settings for downstream access. (There's no explanation why satellite reception speed would vary among these areas.)
Service will be relatively expensive, but with high costs (satellite service, for instance) and high utility (equivalent to in-flight utility), it will be worthwhile to a large class of passenger. Rates are C$8.95 for 24 hours of access; the full run of Quebec City to Windsor is roughly 14 hours. Unlimited monthly service is just C$46 per month, a good deal for a couple of monthly round-trips. As-you-go rates are insanely high: C$3.99 for 15 minutes and C$0.30 per minute thereafter!
This article says it's the first North American broadband train service, which is incorrect. Altamount Commuter Express (ACE) has been running a slow broadband service for years using the same technology provider that powers VIA Rail.
The Wi-Fi service costs $8.95 for 24 hours of access, or $46 for a monthly account. Short-term hookups are available for $3.99 for 15 minutes, plus 30 cents a minute.
A pair of bills introduced in Congress last week want to leverage unused television channels: The two bills want to move forward on allowing wireless broadband over television channels in areas in which stations aren't broadcasting. The New America Foundation, which promotes multiple uses of existing frequencies and open spectrum policies, says 40 to 80 percent of TV spectrum is empty in rural areas. The bills differ in how much of this spectrum they'd allow to be used. When the digital television transition is complete, now mandated for Feb. 2009, the remaining analog frequencies will be auctioned off, and thus if a pre-existing "white space" use were in place, that might reduce the spectrum's nationwide value.
CSR, a major Bluetooth chipmaker, backs WiMedia Alliance's flavor of ultrawideband (UWB): The jockeying as the UWB war nears market entry is underway. CSR is a member of the WiMedia Alliance, but this announcement nonetheless signals their commitment to move Bluetooth profiles--like object exchange and hands-free audio--to work over the Multi-Band OFDM flavor of UWB.
Meanwhile, Belkin told me a few days ago that their wireless USB 2.0 UWB-based dongle and hub combination, relying on a design from Freescale, will ship in July not early spring as originally announced.
Months after a deal was rumored, Fiberlink adds T-Mobile's hotspots to its aggregated roaming system: The company announced the deal yesterday, which brings 6,700 hotspots to the mumble-mumble total that Fiberlink offers. A company representative said the firm declined to get into the raw numbers game, partly because many hotspots in the largest networks--such as Sprint, Boingo, or iPass--tend to mass in specific countries like South Korea and Japan.
Fiberlink uses Boingo's aggregation software platform, the spokesperson confirmed. The deal was first rumored last September. iPass, a key competitor of Fiberlink with its recently completed acquisition of GoRemote, signed their deal with T-Mobile over two years ago.
David Pogue writes about three of the increasingly large number of cellular data to Wi-Fi gateways: He praises the notion of pairing the ubiquitousness of 2.5G and 3G cell data networks in the US with the affordability and utility of Wi-Fi adapters found in every device. He notes how a mobile router that combines WWAN and WLAN (and Ethernet LAN in Junxion and Top Global's boxes) is ideal for traveling workgroups. The Kyocera gateway works with just Verizon and Sprint's cards; the other two handle a wide range.
Pogue notes that despite protestations from companies like Verizon over the inadvisability of violating one's terms of service by sharing a 3G connection, it's inevitable that this kind of sharing will simply be part of the contract. He writes a propos of DSL and cable firms wanting to charge per computer on a home broadband network, "In the end, of course, common sense won, the cable companies lost, and now just about every home D.S.L. or cable modem signal is shared among two or more computers."
Die Zeit offers this report on user-spread Wi-Fi hotspot builder Fon (in German): It's interesting to read about Fon from a European perspective in which there's more optimism about models that challenge incumbents. Perhaps we're too beaten down in the States? I'm quoted on page 2 talking about the hurdles of convincing telecom giants to sign onto Fon, but that secondary ISPs could use it as an effective instrument.
I've thought for years that ISPs like Speakeasy that openly promote sharing would eventually steal higher-priced, better-margin customers from ISPs like Verizon that forbid it. However, the variety of legislation, regulation, and court decisions that affect the DSL and cable markets have prevented effective wireline competition for consumers.
The local business journal reports the city council wants to review the public/private partnership: The plan calls for a city franchised fiber/Wi-Fi network in which the city contributes no capital, but provides telecom business to the winner provider. A local Minneapolis group has pushed the notion that the city should own and operate the network itself, although that group isn't mentioned here. The council has delayed acting on the finalists' bids for the network to review whether that course of action makes sense.
You can hear from several sides of this issue in a December Minnesota Public Radio report that includes Becca Vargo Daggett of the Institute for Self-Reliance, the local president of Qwest, and Minneapolis's Bill Beck, deputy CIO.
Update on Feb. 24: The council decides against ownership, and moves to authorize pilot projects by EarthLink and US Internet.
Dana Spiegel covers the release of a new request for proposals for parks in New York City: The first proposal, released two years ago, covered 17 parks. Wi-Fi Salon won that bid and apparently has just a single park deployed; their model was to find sponsorship. The new RFP covers six parks, mostly in Brooklyn, and requires the winner to offer free access. But Spiegel, the head of community group nycwireless, notes that a franchise fee of some sort is required and a $700 application fee among other requirements or parameters he finds fault with.
Strix gear will be used to support 10,000 voice lines in Cittagong, Bangladesh: An eight-square-mile area will be covered with 90 Strix nodes installed by Accatel and Nextel Telecom of Bangladesh during stage 1 of deployment. The network will ultimately offer 200,000 voice lines over three years. The service will include voice, video, and data.
The head of the university has the right motives, but the wrong science: Lakehead University president Fred Gilbert should probably consult with the physics department (and any epidemiologists on staff) before making a decision like this in the future. He cites a variety of studies that involve enormously higher and more focused electromagnetic radiation and which still produce no conclusive results that point to elevated risks of cancer or other diseases.
I support his notion of prudence, but unless he bans cell phones and nearby cell towers, cordless 2.4 GHz phones, and other microwave emitters that produce higher amounts of microwave output, it's a little silly to focus on the low-power, diffuse, hard-to-tell-apart-from-noise nature of Wi-Fi.
It's hard to ridicule the guy, though, because he's not asserting anything factually incorrect nor is he grandstanding. It's apparent he has the best interests of students and staff at heart, and that should only be commended. [Link via BoingBoing]
Update: I take back the part about not him not asserting factually incorrect statements. Reuters quotes his statement as being, "Some studies have indicated that there are links to carcinogenetic occurrences in animals, including humans, that are related to energy fields associated with wireless hotspots, whether those hotspots are transmissions lines, whether they're outlets, plasma screens, or microwave ovens that leak."
This is extremely sloppy: wireless hotspots aren't a large category that includes transmission lines, power outlets, screen, and microwave ovens. The president's background is in biology and zoology, not comparative literature or interpretative dance, and thus one might expect that a rigorous scientific grounding would inform his public statements.
That's a bad trope on microwave ovens, by the way. They produce a high-intensity beam that dissipates very quickly; it's only reflection within the oven that causes water molecules to twist, heating the food. Thus even a large leak would only have an effect if you were very close, and then very little (unless it was a massive hole) because the beam is moving to produce even food coverage. Can someone teach the rest of the world the inverse square rule?
Sen. Hillary Clinton supported grants for a nonprofit to extend broadband into South Bronx: In a speech on site, the senator suggested that in areas that private firms find it unprofitable to install and operate broadband networks that communities could create their own networks. She backs a federal grant for one such network in the South Bronx that would bring in fiber and Wi-Fi.
While it might seem unlikely that the South Bronx would suddenly turn into a high-tech haven, stranger things have happened. An affordable local work force coupled with low real-estate prices and fiber optic lines has turned many rural and urban blighted communities into enclaves of good-paying jobs. I saw this in Maine, where service jobs started popping up along fiber routes.
Larry Brilliant, founder of AerZone and Cometa, will lead Google philanthopic foundation: Google's $1 billion charitable foundation will tackle poverty, the environment, and energy issues. I note Brilliant's appointment because while a medical expert with a smallpox specialty, he's been involved in two of the most significant early Wi-Fi networks.
AerZone, a division of SoftNet, had contracts signed with several airlines by late 2000, when the parent company abruptly halted operations before networks were launched. Capital had apparently dried up just before the true dotcom crash. Cometa, any regular reader of this site knows, was an ambitious, probably overweening attempt to build the country's largest hotspot network of premium venues but suffered from numerous execution errors. The USA Today article linked to above provides the wrong shutdown date; it was May 2004, not 2003. AerZone was ahead of its time; Cometa, in some ways, behind it. (The article also repeats that AT&T, Intel, and IBM created Cometa as a joint venture. In fact, each firm made investments of cash or services, while two VCs provided other money.)
Brilliant left Cometa long before its shutdown because he felt he couldn't act as CEO and be on call for virus outbreaks. His new job melds the two roles. Brilliant may have faced business model criticism in the Wi-Fi world, but he should receive (and has received) international praise for his roles in improving the health of impoverished people. In his role at Google.org, I expect a repeat of the Seva Foundation--this article says the foundation estimates 2 million people had their eyesight restored via its programs--rather than Cometa.
In a move that doesn't surprise me, EarthLink and Google have combined their proposal to build a network in San Francisco: Both companies have openly spoken for months about how they were working together to understand how viable models for a variety of metro-scale networks could operate profitably for the ISP. Google has said rather consistently as well that they aren't interested in running Wi-Fi networks, but rather need some testbeds to try out ideas--hence their SF and Mountain View interests.
Don Berryman of EarthLink stated in a brief release (also found on the company's Earthling blog), "We will be able to offer services to different customers on the network that fit with their own individual needs and wants." Based on my research and interviews, I suspect this means that Google can offer ad-sponsored free access and perform the tests on efficacy that they want. They can also introduce tests of new services or variations on services, such as their instant messaging and virtual private network software. EarthLink can charge for higher-speed access and provide business-grade service using their Motorola Canopy backbone. (The free/fee tiered basis is now being reported at GigaOm.)
This is another nice win for Tropos and Canopy, by the way, as EarthLink has committed to using gear from those two firms for at least the first five cities that they build out. Other bidders now face a higher bar given the integration of the two firms' proposals.
My friend and colleague Om Malik has been beating a drum for months that Google might build Wi-Fi networks from coast to coast. I have disputed this not based on evidence--Om uncovered fiber optic line leases and other deals--but on margins. Google has a very high-margin advertising business, while ISP service involves ugly amounts of physical infrastructure investment and relatively low margins that require huge volumes to achieve good returns.
Now the EarthLink-Google bid linkage in San Francisco is fascinating as Google originally was to build their network using wireless infrastructure builder WFI. The status of that wasn't mentioned in the brief statement I received from EarthLink, of course.
Update: The San Jose Mercury reports bids have also been received from MetroFi, Communication Bridge Global, NextWLAN, Razortooth Communications and SF Metro Connect (SeaKay, Cisco Systems and IBM). Cisco has new mesh equipment and IBM has a service operation that handles installations of networks.
Most municipal networks that are being bid out combine several kinds of networks and several kinds of services: The coverage and criticism of municipally authorized and bid networks often focuses strictly on Wi-Fi. That's just one element. Most municipal networks comprise six separate components.
Wi-Fi for mobility. This is the typical form of outdoor Wi-Fi that's well characterized and currently employed in networks as large as several hundred scale miles. This is also the form in greatest competition with cellular 3G because outdoor Wi-Fi typically offers much greater upload and some greater download speeds unless heavily loaded with users. Wi-Fi can be cheaply deployed in rural areas--think the giant networks in Oklahoma, New Mexico, and Oregon--where 3G isn't a sensible investment for cell carriers. In urban environments, mobile Wi-Fi will work, but may have competition and interference.
Wi-Fi for residences. This is the riskiest part of metro-scale wireless: having signals strong enough from the least number of outdoor Wi-Fi mesh nodes or access points--least number to keep costs at their lowest--to reach weakly enough into single-floor and multi-floor residences that they don't interfere (much) with existing networks and yet can be picked up with high-gain adapters or bridges.
Intra-node networks. All mesh networks use some form of communication among nodes to achieve the mesh advantage. This type of communications can be in the same band and channel (Tropos, one kind of BelAir router) or in 5 GHz (more common). The former saves cost; the latter reduces contention, reuses spectrum, and has a greater backhaul pool available, but requires one 5 GHz radio for every 2.4 GHz radio, keeping costs much higher.
Back-haul networks. Moving data from mesh clusters into higher-capacity point-to-multipoint connections is a clearly different form of network. WiMax should play a big part here, as it will in EarthLink deployments in which Motorola Canopy base stations--which are very WiMax like and will eventually go through certification—form the aggregated backhaul. These networks are subversive: In most municipal deployments, cities will turn over intra-building data networking to the muni-scale provider who will use WiMax or fiber to link those facilities together. Philly estimates 300 to 500 buildings will add better networking or replace wired leased lines with Canopy connections.
Aggregation. The back-haul networking will reach aggregation points which will comprise either or both fiber optic lines and high-frequency, short-range, very high speed microwave links.
Backbone. The network operation center where traffic is exchanged across the network, to local servers, and to peering points that join to the Internet.
Now the first four functions can be in a single device, of course. But it's good to think about all this entails.
The RFP released from Houston, Texas, today for a citywide network defines these into three categories: public service (municipal non-emergency purposes), public access, and public safety.
Most of the above are on the plate for every single serious urban municipal broadband RFP, which is one reason why fiber--despite its expense--is often cited as a reasonable part of a broadband plan. If not to the home or node, at least a fiber ring to deal with local backhaul and municipal needs.
The FCC seems to have quietly approved the rules necessary to allow 11 more 802.11a channels to the 5 gigahertz (GHz) band: Joanie Wexler of Network World reports the 255 megahertz (MHz) chunk added to unlicensed 5 GHz use comes as a result of compromise between the Department of Defense and the "industry," which wasn't well defined. Public process doesn't seem to have been broadly involved here, but it may just be less transparent to someone not trained in the intricacies of FCC interaction with other parts of government.
The new rules went into effect Jan. 20, and allow the use of 5.47 to 5.725 GHz--11 channels in the 802.11a version of Wi-Fi--with a couple of key signal usage modifiers. The rule also changes the requirements for the next-down stretch of spectrum, 5.25 to 5.35 GHz to conform to those modifiers.
(Update: It turns out this story is incorrect: the 5.47 to 5.2725 GHz band was available about a year ago. This new order adds the signal modifiers below to the 5.25 to 5.35 GHz band. There was a lag of some kind between these new channels being available for legal use and chips and equipment supporting them, but no regulation gap.)
Both bands must now use dynamic frequency selection (DFS) to avoid in-use spectrum, and transmit power control (TPC), which throttles power to the minimum necessary for given communication. Older equipment using 5.25 to 5.35 GHz is exempt, although one might expect manufacturers to push out firmware upgrades. These requirements are part of 802.11h, which extended 802.11a for legal European operation.
Interest in 5 GHz has increased in the last year or so with the recognition that having no 802.11b devices to support in that band makes voice over IP over WLAN (VoWLAN) easier to operate in corporations. The band is also being used for backhaul on mesh networks and will be one of the early profiles for WiMax.
Even though 5 GHz has restrictions on omnidirectional signal strength that makes it propagate less far than the highest-powered 2.4 GHz omni-driven gateways, the rules for directional antennas for the top four channels (highest frequencies) in the band give it parity. The lack of competing signals and now a total of 23 channels (some with limitations on outdoor vs. indoor use) make 5 GHz very appealing for these specific applications.
Chicago is starting plans to have a Wi-Fi network built across its nearly 230 square miles: The city's CIO said companies will be asked to talk about how they might build the network, which sounds like an RFI (request for information) rather than a more concrete RFP (request for proposals). The CIO said quite firmly that no city money would be involved in building or operating the metro-scale network.
Oddly, Steven Titch of the Heartland Institute, a perennial opponent to the construction of municipal telecom, is quoted as being in favor of this approach. It's possible there's a subtlety missing in his quote, or that the institute has moved to supporting cities that are essentially franchising the building of wireless networks. But this jibes with the statement earlier this week by the CTIA, the cell industry trade group, that it's not opposed to these kinds of networks, either, as their members might bid on them. Update: Titch confirms (see comments below) that his views are accurately represented and notes there's no connection with the CTIA statement. He also points readers to reason.org for more on the franchised city-wide network topic.
One commentator noted that if you stop calling a municipally requested network a "municipal network" and call it something else, like metro-scale or city-wide, you suddenly remove the notion that the city is controlling or funding it. That semantic change may be significant.
The RFI or RFP will be issued in spring with recommendations for the mayor and city council before fall.
The Bluetooth SIG adds Phone Book Access Profile (PBAP) for better mobile device/car integration: This new profile allows a mobile phone or other Bluetooth-supporting device to connect to automotive telematics (on-board electronics) that have similar support and gain access to the device's phone book, status, and features like hold, call waiting, and caller ID. It provides enhanced audio, too. This support will probably be found in cars in early 2007.
The National Journal reports the CTIA says it's not opposed to competition from municipal broadband: Many of its members offer Wi-Fi service and Cingular apparently bid on the San Francisco service, which I'm embarassed to have missed. (Cingular is majority owned by AT&T, minority owned by BellSouth. AT&T has led significant lobbying efforts in Texas against municipally owned or franchised broadband, but Verizon and Comcast have been loudest elsewhere.)
Update: A colleague wrote in to note that Cingular's "bid" in San Francisco was a response to the initial RFI/P (request for information/proposal) in which they noted the prevalence of 3G service. I tried to look up their response and it's not in the archives at the Tech Connect site that houses SF's project.
Further Update: Another colleague wrote to tell me that Cingular's "bid" was part of the SBC response to the RFI which is not found in the alphabetical proposal listings, but rather in the Comments download, which was to be reserved for non-proposal input. On reviewing that submission, it's clear that SBC answered with a great degree of detail and no real SF specificity except to state that 73 FreedomLink hotspots are in SF. They talk at length about SBC's fiber-to-the-home (FTTH) and fiber-to-the-node (FTTN) initiatives without noting how many SF residences might benefit. They mention 3G service being ubiquitous without discussing monthly costs or interior access (poor). They suggest DSL service up to 6 Mbps is available, but don't have any information on which fast speeds are available to which percentage of residents. Hardly a bid for Wi-Fi service.
The first summit hosted by CUWiN was in 2004, and apparently connected a lot of dots (and faces) for those focused on politics and others on technology: This year's event, March 31-April 2 in St. Charles, Missouri, splits tracks for technology, implementation, and policy. Based on the previous event, there's very little talking heads up on a stage, and more symposium-like interaction. The cost ranges from $65 to $675 depending on type of registrant.
The name Fon means more than a trademarkable version of a common household and handheld appliance: In an interview this morning, founder Martin Varsavsky explained that Fon's focus isn't on laptops and plain handhelds. Rather, Fon's focus is on cellphones, game consoles, and cameras that have Wi-Fi built in--or that soon will.
Fon is a grassroots hotspot network in which downloading firmware and installing it--and no cost--onto an appropriate Wi-Fi gateway device turns the gateway into a Fon hotspot. I spent a chunk of last week writing and being interviewed about Fon's potential to build a network of what they call Foneros. I spoke to Varsavsky about issues I saw as red flags and others raised by readers of this site and bloggers worldwide.
The executive summary is that Varsavsky has focused Fon on mobile devices, not laptops, and that the company will try to increase the number of Internet connections that ISPs sell while simultaneously increasing the revenue they derive from each connection. Mesh and community-spanning networks are part of future thought, but not announced plans. And Varsavsky acknowledges that his attempt to be informal in announcing the investment may have backfired: too much was read into everything he wrote (and has written for months) and that his US advisory board wrote. He was just trying to bypass the usual press release and stiff statements that accompany funding announcements.
I'm still convinced that Fon faces high hurdles in achieving the kind of market they're longing to create, but I'm no longer confused as to the kind of network they're building. There are many aspects to the nuances of what Fon could involve that may give it a foothold. Many parts of Fon are similar to ideas held by community wireless networking groups today--they differ on who owns the physical part of the network, but in terms of local services that run over a local network independent of the Internet, Varsavsky's on the same page. Now on to the detail.
Varsavsky said that Fon isn't per se unique, although it's sometimes represented in that fashion. He was aware of previous efforts, such as Sputnik's original business plan--he's a friend and colleague of my friend and colleague David Sifry, a founder of Sputnik and the founder of Technorati. "I don't really know how original we are, but I know we are timely," Varsavsky said, noting that of seven companies he founded and six he sold he had just one he considered a financial failure--a German ISP--but only by selling it a couple of years before the market was ripe.
He believes that Fon is different than previous efforts because of the focus on a simple hardware device rather than (as with Sputnik, Joltage, and SOHOWireless) a PC with an appropriate network card. "The way I define Fon is a software download that turns your router into a global family of routers that shares Wi-Fi," he said. "We centered on the router and we concluded that people wouldn't want to leave their PCs on all the time."
Some free and fee competitors also use authentication (via the 802.1X protocol or a gateway page), but those solutions don't extend into the hardware gateway to provide bandwidth throttling and other tools. The open-source packages that Fon uses are available in various forms for standalone gateways, including something close to what they provide today. But these open-source projecdts don't have unified account authentication behind them as that requires services. Systems like Radiuz, in other words, don't reach into the network layer of the router while open-source packages don't reach up into the application and access layers.
Varsavsky knows that some of the early criticism has been directed at the idea of average users flashing--or changing the firmware--of Wi-Fi gateways. He points to the iPod to refute this, interestingly. iPods are frequently flashed with new firmware via installers provided by Apple. "When people said how is this? We said, do you have an iPod, it's just like that. The iPod trend as worked for us," he said.
Where Varsavsky says Fon will make its impact is via the tens of millions (or more) people who will be carrying around Wi-Fi equipped cell phones, game-playing systems, and cameras in the coming years. There are already at least eight million Nintendo DS units and Sony PlayStation Portables (PSP) sold in the US with Wi-Fi included. Many millions more have been sold internationally.
Varsavsky also points to Kodak EasyShare-One camera, which uses Wi-Fi to transfer photos (although I find the camera needs additional features to make it a true Internet camera rather than a Kodak service camera), and cell phones with Wi-Fi. At the 3GSM conference in Barcelona, Spain, today, Nokia introduced the Nokia 6136, which allows true seamless roaming between GSM and Wi-Fi connections. Phones announced before this required calls to be placed or received on either form of network and remain on that network for the duration of the call. "That's a fantastic trend, and a trend of all the Wi-Fi enabled phones and PDAs," he said.
This focus on mobile devices rather than computing devices makes the investment by Google, Skype, Sequoia, and Index more explicable. Laptop hotspot venues largely already have Wi-Fi access, are in the process of being signed, or are disinterested. It's hard to find locations that simply aren't locked up or unavailable. But mobile devices can benefit from a different kind of geographic availability of Wi-Fi, and typically require less actual throughput ot be useful. (Cameras are a particular category: upstream bandwidth will be a high-demand item for cameras, especially as they add megapixels. And Wi-Fi hotspots with higher-speed upstream links than home asymmetrical service might be in greater demand for that segment.)
Follow the link for the rest of this interview...
Esme Vos receives a flattering portrayal in the Wall Street Journal: As anyone who reads this site knows, Muniwireless.com evolved quickly from a grab-bag of details to the best source of information on municipal networks. Esme has a bias and doesn't hide it, nor does she makes any bones about advertising and investments from firms with an interest in the success of municipal networks. But what makes her advocacy useful is that she has the facts to back them up. Her reports on proposals and deployments of municipal networks eclipses other information in the field. I'm sure that those who oppose municipal broadband--for competitive or ideological reasons--consult her site frequently as well.
I'll be talking to Fon's founder, Martin Varsavsky, tomorrow morning: After a week in which I expressed my opinion in my own and other forums while battling deadlines that nearly did me in, I'll finally have a chance to talk to Varsavsky, who is making time to speak while on what sounds like a very rare vacation.
I've already got piles of questions for him about Fon's relationship to ISPs, the potential revenue that an individual hotspot might achieve, and how Fon will work in different countries. I'm also curious about the number of router models they'll support, and whether mesh routing software and larger antennas are part of the way in which Fon's deployers (Foneros) will be encouraged to extend access. (Mesh would seem to be in conflict with the one Internet connection, one Fon hotspot model, but we'll see.)
If you've got a question for Varsavsky that you'd like me to pose about Fon, please post that question in the comments below, or email me with them. I can't promise to ask or get an answer for every question--startups like to keep their future plans close to the vest, naturally--but it'll be good all around to hear what other people wonder about when looking at Fon's publicly announced plans.
From my point of view, I'm hoping the furor around Fon dies down so that we can look at the broader scope of their aims. Read Ethan Zuckerman's post on Fon, for instance, from last week, where he talks about his role in the advisory panel in context of extending access in Africa and the developing world. He notes that the incumbent telecommunications companies there may be hostile to reduced revenue from networks that rely on very few paid injection points, while sharing a paid connection over a short distance and receiving a revenue share on that might be highly palatable. Zuckerman writes, "I've looked closely at projects designed to build community wireless networks and have been frustrated that many of these projects seem designed explicitly for nations where bandwidth is cheap."
The WiMedia Alliance promises interoperability among members' implementation of the MultiBand OFDM version of ultrawideband (UWB): Their first set of testing resulted in Alereon, Staccato, RealTek, WiQuest, and Wisair's silicon working nicely with each other. The group expects devices on the market in the second half of 2006. From conversations I've had, I expected that Wireless USB will be the first real application. But the way in which the WiMedia forum wants its technology implemented--one radio, many standards--early radios could later support overlays of Bluetooth applications and FireWire transfers.
Certification for the PHY (radio/physical) part of the equation should start by the end of this quarter; higher layers, like MAC (data framing and error correction) and applications will start being certified by the end of 2nd quarter.
The alliance's chief competitor, Freescale, has devices due out in in a month or so, initially USB replacements from Gefen and Belkin, two smaller but significant niche networking, video, and accessory firms. The first USB hub with UWB will be a paired set of a dongle and an AC-powered four-port USB hub that can be located some distance from the laptop. The devices will use UWB for communication but simulate USB 2.0 for the ports, requiring no drivers.
Freescale has discussed UWB has a seamless replacement technology in its early products rather than as a multi-purpose radio with various overlays. However, that, too, is to come from them, as the Bluetooth SIG has publicly--and one images other privately--have discussed adding their higher-layer services on top of Freescale's flavors.
IBM is testing a chip that will conform to a new IEEE standard for high-speed, short-range networking: This 630 Mbps product that they've tested in the lab conforms to 802.15.3c, a standard that most of us haven't been exposed to yet. Recently, the 802.15.3a group voted to disband; they focused on ultrawideband (UWB).
In the IEEE, every letter and number counts. The 802.15.3a group is Working Group 15 for personal area networking, Task Group 3a for high-data rate PHY (physical layer covering radio technology). This group was deadlocked between the WiMedia Alliance proposal led by Intel and Freescale's classic UWB technology.
Now Task Group 3c is a different story. Task Group 3 (802.15.3) defined a new MAC (framing data, error correction, and so on) along with a PHY that was regarded. The 3c task group builds on that MAC but focuses on millimeter-wave (mm-wave) microwave for extremely high rates. The draft standard is due by November, with a final version expected Sept. 2007, this article notes.
The IBM technology uses mm-wave in the unlicensed 60 GHz band (I honestly didn't know such a band existed until today) to communicate in this first draft at 630 Mbps over 10 meters. Some forms of UWB are nominally expected to deliver 480 Mbps at 10 meter, although some forms already available in sampling have higher rates. IBM thinks that 630 Mbps is just the beginning, and Task Group 3c is looking at rates that could exceed 2 Gbps.
Connexion scores another 15 planes, with potentially more: The leading People's Republic of China carrier committed earlier this week to install Connexion's in-flight broadband service on as many as 15 planes initially. Other planes may be encompassed in the deal later. The initial planes are all 747-400s. Installation will begin in October. The goal, of course, is to have this advanced service in place before the 2008 Olympics in Beijing.
John Cooper has launched MetroNetIQ: The well-designed site draws together many different elements of the evolving metropolitan-scale and municipal networking world, including links to Web sites, articles, books, and resources.
Madison starts its unwiring: The downtown and nearby university will be covered. Service will be free for at least a month when fully operational. It's being built as a vendor-neutral wholesale network by Cellnet.
Voice, video, and data will cover 140 square miles of Nassau: The service will be deployed by Sawtel using Meru gear, which has a VoIP focus. Behind the Wi-Fi coverage, there will be centralized phone switch options for businesses, too.
Boston tries take two: The mayor convenes a task force to focus on putting more wireless in Boston. Last year's effort with a Wireless Summit produced conversation, but no plan for action.
Cincinnati lets Bells put Wi-Fi in libraries: Some libraries find funds tight, and in those cases, have turned to for-fee providers. Cincinnati's system will have Wi-Fi available, some starting this week, at $4.95 per hour or $9.95 per day. However, subscribers to operator Cincinnati Bell's existing ZoomTown and FUSE customers pay nothing extra.
After cogitating on it for a couple of days and talking to other reporters, it's pretty clear that Fon has a great smokescreen with its free "Linus" model: The fact is, Fon is just a paid hotspot operator that wants to eschew all venue signing, installation, and management costs. Fon has three kinds of hotspot users, but attention has focused on the Linuses, those who opt to let others who are also Linuses use their hotspots for free. This will always be a subset of all users. Let's say Fon reaches 1m installed hotspots by 2010, one of the stated goals of the founder. If even 750,000 are "Linus" hotspots, only 750,000 people will be able to use those hotspots at no cost. Six billion people--okay, the tens of millions of travelers with Wi-Fi laptops or handhelds or phones--will pay.
Let's not lose sight of the fact that there are now hundreds of aggressive hotspot builders cherrypicking the best remaining locations. If Fon only gets the dregs of locations (suburbs, apartment buildings, coffeeshops without service right now), then they're a niche-filling for-fee network. If they offer roaming across other networks and resell to aggregators, then they become a backfill network for missing locations, but only if people in those locations find it worthwhile to collect small fees in exchange for sharing bandwidth.
Primarily, Fon is for-fee. Primarily, it faces the same location, location, location issues as other for-fee operators. Locations that want to offer free Wi-Fi will not install Fon.
Update: Esme Vos has a quasi-rant, too, on this subject in which she wonders precisely how Fon fits in the current ecosystem. She suggests that sharing a connection (or having an ISP sell you a device that shares your bandwidth) produces uncertain amounts of bandwidth for your own users. (A FAQ at Fon does mention that bandwidth throttling is a future planned option for the software.) And, she asks, isn't municipal scale better than house to house?
Fremont, Calif., may have Wi-Fi metro-scale provider within two months: I'll take a stab and say (without any inside knowledge) that MetroFi is the undisclosed company that wants to install service at no cost to the city. MetroFi covers Cupertino, Sunnyvale, and Santa Clara in the southwest bay. Fremont is pure east bay (and my former home as a youth), and would be a good high-tech area to add. Fremont is treading carefully: the local utility threatened to sue them when Metricom went under and left their radios on poles! I never heard about that threat. Fremont went out and removed the radios, worthless once unmounted.
Speakeasy says that Fon claims a deal where none exists: I just received this missive and permission to share it from Lynn Brackpool, a spokesperson for Speakeasy Networks. Fon said in reports to some outlets yesterday that Speakeasy Networks was a partner in its plan to spread Wi-Fi sharing using special firmware for Wi-Fi gateways to create massive, grass-roots networks. Speakeasy says they're not involved. (See the bottom of this post for the resolution.)
"FON has falsely announced that they have reached an agreement with Speakeasy regarding support of FON's wireless product. This is 100% completely false. No relationship, financial or otherwise, exists between Speakeasy and FON. In fact, Speakeasy was the first to pro-actively support wireless broadband sharing amongst its subscribers many years ago. Indeed, FON's product appears to be built on a premise Speakeasy introduced in 2003 called NetShare which involves individual revenue-sharing agreements among wireless users (http://www.speakeasy.net/press/pr/pr070803.php).
"We are currently involving legal counsel to demand a retraction from FON regarding their misleading announcement."
On Fon founder Martin Varsavsky's blog, he wrote, "Also I am pleased to announced today that we have obtained the support of two significant ISPs for FON. In America Speakeasy has said that they welcome FON and in Europe, Glocalnet and FON have signed an agreement so Glocalnet sells its services FON ready and the Swedish foneros will soon be able move around Stockholm and other cities with their WiFi enabled gadgets."
This isn't so much a statement of partnership, but might reflect Varsavsky's notion that Speakeasy wouldn't block Fon users.
On the other hand, he several times in his blog and in news stories mentions the word agreement, support, bargain, revenue sharing. Speakeasy has no agreement of any kind with Fon, which would tend to contradict any sense that Fon was sharing revenue with them (unilaterally?) and thus argues that Varsavsky was trying to broaden his appeal by mentioning a U.S. ISP.
Question for Google, Skype, Sequoia, and Index: Did Varsavsky claim a Speakeasy contract? If so, did you do due diligence? If not, will he disclaim his statements?
Update: Speakeasy's Brackpool sent this follow-up statement late today:
"No relationship, financial or otherwise, exists between Speakeasy and FON. The impression may have been created by FON because Speakeasy has always supported an open wireless sharing policy. We have contacted FON and they too understand that no implicit agreement exists between Speakeasy and them in terms of their service. We are satisfied with their response and do not plan to pursue any legal recourse."
I have noticeably ridiculed Fon since its hapless announcement and subsequent hype fest two months ago (Reuters, Associated Press, New York Times, Wall Street Journal, Fortune): Today, news broke that instead of a bottom-up, grassroots-driven effort to spread Wi-Fi access through individuals installing Fon's firmware on commodity gateways, this group is aiming to become a multi-hundred-million-dollar company. That's more like it.
Google, Skype, Sequoia Capital, and Index Ventures have put in nearly $22 million to spread Fon's idea, which I suspect will involve Fon installing seed nodes through the world. Fon's model, partly implemented, involves three categories of users: free operators, for-fee operators, and non-operators. The free operators, or Linuses, provide no-cost access to other Linuses; for-fee operators, or Bills, charge everyone and receive no free access; non-operators or Aliens pay for service at all locations.
Part of why I'd ridiculed Fon was because it seemed to rely on an idea already discarded in the Wi-Fi hotspot world that by providing the right software and incentives, tens of thousands of people would install hotspots, building a worthwhile network. But this is problematic because hardware and physical locality are tough sells.
Skype is useful when even two people have Skype in the entire world. Those two people, if geographically distant, save money using it and have better quality calls. With a few thousand people on Skype, it becomes viral because its value is much higher the more connections among people. A few million, and you have a multi-billion-dollar acquisition by eBay.
Fon is harder because home gateways of the type they are providing software for initially--some Linksys models with Linux embedded operating system software--can't produce much of a signal. Home users can provide service to relatively few other home users via these gateways. There's only utility in the Fon network if you're charging for service and a lot of people use a hotspot, or if you're not charging for it and a lot of Fon locations exist in the places you travel to (around town or the world) for you to also use for free.
The investment capital they've raised signals that while their network might be extended on a peer-to-peer basis, they'll be putting their money into seeding the network by putting nodes in well-trafficked areas, probably in the thousands.
They have a real ISP problem, discussed extensively in the AP and Reuters story on Fon's incoming capital, because (as the AP reporter quotes me saying) it's trivial for ISPs to track Fon operators if the ISP bans sharing an Internet connection. Fon has central authentication so even with encryption, the destination of Fon messages will be known. (Of course, if Fon uses a peer-to-peer model like Skype, they could tunnel authentication through peers making this impossible, but that's not a reliable way to run a login system.) Speakeasy and Sweden's Glocalnet have signed up, Reuters reports; Speakeasy is the only national ISP I am aware of in the U.S. that encourages sharing their connections. (Update: Speakeasy says there's no deal.)
Om Malik points out that Skype has concerns about being blocked by ISPs, so an alternate network of sorts--ISPs are still in the middle of this--benefits them, while Google is sorting out Wi-Fi business models. I argue strongly that Google will not become a Wi-Fi provider beyond San Francisco and Mountain View (at least not on any large scale) because their interest is high-margin businesses like advertising not low-margin ones like service provision.
Fon can only succeed with critical mass being achieved quickly enough to encourage the grass-roots uptake they need and with ISPs not banning them as critical mass grows. (Fon's founder said they'll subsidize the gateways so people can buy versions with Fon software installed--flashed over existing firmware--for $25, instead of $50 to $70 retail.)
And shouldn't municipal-scale networks with limited free access kind of distort the Fon model? Why use Fon when you can get 300 Kbps for fee (Google's bid for San Francisco) or two hours a day for free (many other cities)? If Google is as successful in building a business model for ISPs to run municipal wireless networks, advertising-supported service--such as what MetroFi now offers in three Bay Area cities--seems to wipe the floor with Fon.
Lest Fon be seen as entirely new and unique exclusive of the three firms mentioned before that tried this model using cheap computers and software instead of commodity gateways and firmware--Joltage, SOHONetworks, and Sputnik--remember that LessNetworks and Radiuz have been offering community authentication for a while. LessNetworks has been kicking around for a couple of years, a for-profit outgrowth of the Austin Wireless City Project that offers a gateway page, community features, and user accounts for free access for a small management fee; they donate their services to nonprofits and others, too. Radiuz is free WPA Enterprise authentication, which combines secure logins and unique WPA encryption keys with RADIUS features found in virtually all consumer and enterprise gateways. Radiuz started up a year ago.
Update: I must be naive, but thought of this as a relatively small story. I am now seeing separate accounts in practically every major newspaper and publication in the world--I suppose Skype (eBay) and Google will do that to a company. Read what Doc Searls has to say about it, and also David Weinberger, who is part of Fon's US advisory council.
Update: Fortune's article mentions that the Fon founder brought on board to that US advisory council a whole array of digerati and A-List technology bloggers.
I know, I know, I said I wasn't going to engage these folks again: But it's hard not to. Joseph Bast, head of the Heartland Institute, is now a martyr because people send him unpleasant email. In his latest foray into shaping public opinion, Mr. Bast was given an op-ed column to write about the failure of municipal networks in the Fort-Worth Star-Telegram. As usual, Mr. Bast uses outdated information (despite his statement that some numbers are new), poor analysis, selectivity (many cities he used to cite are missing because their "failures" were disproven, notably Tacoma, Wash.), and the bogeyman of taxpayer risk.
I never see a statement about taxpayer risk when millions of dollar are provided to companies that move to town, have roads built for them, hire local workers, and then leave when tax giveaways expire, and it's no longer expedient for them to remain, nor contractually required. All money given to corporations through subsidy and tax incentives is positive in the view of ideologues (sponsored by those companies or not); all money spent by cities on economic development and basic services is wasted.
Here's the letter I wrote to the editor of the Star-Telegram. Note that I am not stating that Ron Rizzuto is providing inaccurate information nor is a paid lobbyist. Rather, as with all organizations and individuals engaging in public debate, it is as important to listen to what they say as who they work for. Prof. Rizzuto doesn't hide his affiliation with the cable industry, but those that cite him would prefer to mention the University of Denver or the Daniels School (named for a cable industry pioneer), not his work with The Cable Labs, a well-respected cable-industry funded institution for research, among other tasks.
(Study up on this issue from both sides: read reports at The Heartland Institute, Pacific Research Institute, and The New Millennium Research Council; as well as Muniwireless.com and Free Press [Telco Lies, principally], as well as this older page of refutation of muni failures at Fiber For Our Future.)
Joseph Bast's opinion column stating municipal broadband networks suffers from several factual flaws and omissions, none of which relate to his technical but all to his political abilities (Star-Telegram, Sun., Feb. 2006).
At least one of the reports Mr. Bast refers to that his organization produced was released with The New Millennium Research Council, a wholly owned subsidiary of Issue Dynamics, a Washington, DC, press relations firm. Neither NMRC nor Issue Dynamics hides this fact. Issue Dynamics represents a variety of organizations, including Verizon, AT&T (formerly SBC), and other telephone and cable companies that provide existing broadband services.
The NMRC board members and other organizations that contributed to that report had direct ties with either Issue Dynamics or Verizon when the report was issued. Mr. Bast's Heartland Institute does not disclose its corporate funders.
The networks that Mr. Bast refers to in his column are all fiber optic networks, which require great effort and expense to build and operate because of digging up streets and installing conduits. The enormous and expensive overbuilding of fiber optic networks was one of the leading causes of bankruptcy and instability--along certain alleged accounting actions--of major telecommunications operators at the time of the dotcom crash.
The municipal broadband networks built, in progress, or bid out in nearly 200 U.S. cities at this writing are largely or entirely comprised of wireless networks using standard Wi-Fi for user/citizen access. These networks are often being built in cities and towns that have little or no broadband service and no commitment from incumbents to build such; or, true broadband speeds are available in limited, higher-income parts of town.
The expert Mr. Bast cites, Prof. Ronald Rizzuto, even states in the co-sponsored Heartland/NMRC report, "the breakeven point for Wi-Fi systems is significantly lower than that of wireline networks because of the substantially lower capital and operating costs." Mr. Rizzuto is dubious about other associated factors, but that is the first of his conclusions. (Mr. Rizzuto is a professor at the University of Denver's business school, but also a senior fellow at cable industry funded The Cable Center. He lists former telephone company US West and cable operator TCI among previous consulting clients on his university profile.)
After citing a small number of cities' fiber optic business model flaws, Mr. Bast concludes, "These case studies are important because advocates of municipal broadband systems say they won't cost taxpayers anything."
Unfortunately, Mr. Bast uses the same logic of obfuscation here: not only are the case studies he cites involving systems other than those under consideration by most cities, including Philadelphia's large effort, virtually all current municipal broadband plans require the winner bidder to bear all risk and pay all costs of capitalization and operation.
It is, in fact, much more like a cable franchise agreement in which rights to city facilities are traded for a commitment to provide a service; some cities will even charge municipal-scale providers franchise fees for operation. Others will subject them to local telecommunications taxes.
Mr. Bast and others affiliated with him have fought a multi-year effort ostensibly on behalf of taxpayers and incidentally on behalf of incumbent broadband providers typically by using outdated financial information and citing city networks as failures by employing accounting techniques that are not in accordance with generally accepted public and private measures of return on investment. The Heartland Institute and its associates consistently look only at a limited number of municipal operations, ever changing as their facts are refuted, and never find an iota of success--there is, apparently, no exception that proves their rule.
There are many difficulties political, financial, and technical with building city-wide wireless networks, but few of the real problems have been discussed with academic rigor in open forums by either pro or anti forces. Mr. Bast lacks the interest in writing reports in which both alleged successes, such a Tacoma, Washington, and alleged failures, such as Cedar Falls, Iowa, are presented with independently assessed and audited financial results and returns.
The AP reports that Robert M. McDowell will occupy the third Republican seat on the FCC commission: While he may be pro-something, he's not coming in with a pro-ILEC (incumbent local exchange carrier) position given that he was until this moment the senior vice president and assistant general council at the Competitive Telecommunications Association, a CLEC or non-Bell trade organization.
This is fascinating from the perspective of municipal wireless and other services, because registered CLECs and similar organizations will be among the bidders and beneficiaries of municipally bid or franchised broadband and telecom networks. With two Democrats on the commission generally supporting broader ownership of media and local control, Mr. McDowell could wind up voting against pro-ILEC positions. That remains to be seen.
Red Herring reports Tropos said it's talking to cable firms about using their gear for outdoor Wi-Fi: It's not a surprise. Comcast's investment arm injected cash into BelAir a few weeks ago, which simultaneously announced cable-plant-powered mesh access points. Comcast has been one of the most vocal firms opposing municipal broadband, but most recently offered a year-out-of-date critique of the plans based on cities building the service themselves and funding it from taxpayer dollars.
(This Red Herring article linked says that "cable companies have fought publicly backed city Wi-Fi plans in courts," but I'm unaware of any injunctions, orders, or lawsuits.)
In a recent interview, Tropos's CEO Ron Sege told me in reference to the BelAir investment, "I can give you a long list of captive venture capital arms investing in things that are never used by the operating sides of their companies." This seemed to signal that a Tropos cable deal might eventually be forthcoming. Cable settop box maker Scientific Atlanta was a Tropos partner, but the company was acquired by Cisco last year which now offers its own mesh access points while also owning consumer/SOHO/SMB networking firm Linksys.
But Sege thought that the cable plant wasn't the ideal place to pull capacity from. DOCSIS, the cable industry encoding standard for data over unused local channel spectrum, lack the "upstream capacity," Sege said. He thought there was a large mismatch between even a single 802.11g gateway--capable of at over 20 Mbps of real throughput--and the current cable base. DOCSIS 3.0 has more upstream capability, but it's highly asymmetrical.
In a world where asymmetry is reducing--"broadband is becoming more symmetrical"--Sege said, "It's not clear that the DOCSIS plant is going to be useful in this environment."
Meanwhile, I spoke to BelAir's CEO Bernard Herscovich recently as well, and he said that cable companies' "emergence as wireless players is probably going to be the most exciting news in our industry in the next 12 to 18 months." He said that cable firms have an advantage because "the architecture in the cable space is an IP-oriented approach, very close to what people call NGN or next-generation network architecture."
I had heard from some sources that franchise agreements with cities might limit cable firms deploying mesh nodes on their rights of way, such as poles. Hescovich said no. "They would be able to go and place a mesh node without requiring additional approvals," he said. With BelAir gear, that can be directly powered by the cable plant, "The strand in the cable network has power. There are no power issues, there are no site acquisition issues."
He also said that multiple system operators (MSOs), cable firms that have many operations, have a cost advantage. "Our estimation is that it would be about 30 to 40 percent cheaper by an MSO to build such a network compared to somebody who doesn't have that kind of network," Herscovich said.
Roofnet software will extend service from MIT's campus to pass 100,000 Cambridge, Mass., residents: MIT is extending its network as a town-gown gesture, and initially plans for free access and will gauge response. The network will be built using Roofnet software, one of two open-source mesh-routing protocol projects underway. (The other is at CUWiN.)
This article from MIT's The Tech online newspaper has a number of technical errors, but all of them attributed to various sources, who may have provided incorrect details. For instance, "Philadelphia is creating its own utility, [Cambridge CIO Mary P.] Hart said, while San Francisco has partnered with Google to become a wireless city." As readers of this site know, Philadelphia has a non-profit that will handle relations with EarthLink, hardly its own utility, and San Francisco hasn't chosen a winning bidder yet.
Hart is also paraphrased saying, "Mesh technology allows individual computers to propagate the network and act as new access points, making it unnecessary for a user to be within range of the original wireless signal." This is true of mesh networks in which clients and nodes are using mesh routing hardware and software. For normal Wi-Fi, this is untrue: individual computers connect to nodes which mesh over the same or other frequencies with other nodes.
The Associated Press has a very sketchy version of this story out, too, which cites the same technology error: "MIT researchers will help test the "mesh" technology, which allows individual users to both receive and send a Wi-Fi signal, so that a user doesn't need to be within range of the original signal from MIT."
Wrong, wrong, wrong. (Although see Dana Spiegel's comment on this post.)
Mesh Networks, a company acquired by Motorola (and rebranded as Motomesh), is the only firm I know of that has offered commercial products with individual devices having mesh properties, rather than just infrastructure and backhaul. Motomesh abandoned this plan. Their latest offering, which can contain up to four radios (a combination of 2.4 GHz and 4.9 GHz using their proprietary algorithm and Wi-Fi), doesn't mesh across devices.
Update: Per the comment below, Motomesh does still sell adapters that act as mesh nodes--my mistake! But they're not standard Wi-Fi adapters; they're proprietary products. Their new system doesn't require these proprietary adapters, but it's an option.
Analysts kept pegging the year at which the tables turned as closer and closer, but they were conservative: The numbers for 2005 are in, and 50.9 percent of computers sold in the U.S. at retail were laptops. This does exclude massive corporate and industry purchases, but it's a good reflection of the adoption of laptops by a broader segment of users. News.com notes that base-level laptops cost $500 to $700, while $1,000 purchases a nicely full-feature higher-end system. You can spend a lot more, but that typically buys you battery life, a better screen, and extras rather than radically improved system performance. (Intel Core Duo will change that equation, too.)
A couple of years ago, predictions were that 75 percent of laptops by now would have Wi-Fi as a built-in option. I thought it was nonsense: it would either be close to 100 percent or replaced by another wireless networking technology. The tipping point of cost and utility was already in place when that prediction was made: it just required manufacturers to have enough time to revise their laptop lines to incorporate the technology into current products.