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.
The rise and fall of muni-Fi (and rise again): Clearly, the largest story involving Wi-Fi in 2007 was the at-first continued growth in cities awarding contracts with no money involved on their part to have service providers build Wi-Fi networks--and the subsequent failure of these networks to be built. Starting quietly in late 2006, the market shifted for metro-scale Wi-Fi. During 2007, providers decided that bearing the full cost of a city-wide network without city contracts wasn't financially sensible.
The full scope of the low uptake rates in cities that had large portions of the network built out also became clear: rather than 15 to 35 percent of residents subscribing, just a few percentage points would put a network in the top tier. Revenue is apparently also pretty minimal even in cities like Taipei, Taiwan, the network provider for which was predicting 250,000 subscribers by the end of 2006, and had just 30,000 regular users each month at last public report in early 2007.
MetroFi started to tell cities that without an advance service commitment at a minimum level -- an anchor tenancy -- the company couldn't proceed on networks. In 2007, MetroFi lost half a dozen bids or saw contracts canceled due to this change. Its work in Portland, Ore., the biggest network it was building, won't be extended beyond current limited dimensions until additional capital or a city commitment is obtained; the city has said it won't commit to service fees, however.
Meanwhile, EarthLink lost its CEO Garry Betty in January due to cancer. A strong backer of new initiatives to change EarthLink's core business, his death was certainly one of the causes in a quick re-evaluation of the municipal wireless division. New CEO Rolla Huff pulled EarthLink out of new deals, suspended existing ones, laid off hundreds of employees while gutting the metro Wi-Fi division, and appears poised to leave currently built or underway networks, including their flagship Philadelphia effort. They may sell the division, but it's hard to see much worth in it given the current state.
In a smaller bit of news, Kite Networks, formerly known by various names, was sold by parent MobilePro to Gobility with conditions that according to SEC filings by MobilePro weren't met. Kite was once high flying, in the company of EarthLink and MetroFi as one of the major U.S. Wi-Fi network builders. Now it's still in that company, with work on its Arizona networks apparently halted. A suitor has emerged in the form of a regional telecom that specializes in the Hispanophone market (double entendre intended), and which thinks it could boost Tempe subscriptions from the current several hundred to about 300 times that number. Hope springs eternal.
And while AT&T was able to launch a Riverside, Calif., network with MetroFi handling the installation and operation, it backed out of St. Louis, Mo., due to a utility pole problem, and the bidding in Chicago, too. The Metro Connect consortiums in Sacramento and Silcion Valley were unable to raise financing despite the apparent blue-chip participation by Cisco, IBM, and Intel.
County-wide Wi-Fi was also hit again and again by providers who pulled out--CenturyTel in Pierce County, Wash., for instance--or problems with technology or utility poles. In a few scattered areas, Wi-Fi across counties has been built out, but it's not an idea whose time has yet come.
Muni-Fi isn't down for the count. While these high-profile networks in large cities and county-wide networks have mostly hit the skids, more modest networks with well-defined goals continue to be built with a focus on public safety and municipal uses in hundreds of small and medium-sized towns. Brookline, Mass., may be a good example, in which a public safety/public access network was built relatively quickly and with no reported problems.
And there's one big city success story: Minneapolis, Minn. While local provider US Internet wound up spending more than they'd intended, reports from the ground indicate that service works quite well, and subscriptions and interest are quite high. The company was able to respond almost instantly to the bridge collapse a few months ago by deploying additional mesh infrastructure to add network capacity in the area. And it says that it could reach positive cash flow in early 2008. One of their advantages? They secured a substantial commitment from the city for the services they built.
Other trends of the year gone by: Music and Wi-Fi are clearly more aligned, with the new Zune models and firmware from Microsoft allowing wireless sync (but not yet Wi-Fi purchases), and the introduction of both the Apple iPhone and iTunes touch, which allow music purchases over Wi-Fi but not synchronization. (While the MusicGremlin preceded both the Zune and iPhone/iPod options, it didn't seem to gain any market traction in 2007.)
Security continues to be a concern in 2007, although less of one as home users have clearly accepted WPA Personal, at long last, and networks are increasingly encrypted through better software from major hardware manufacturers. Wizards make encryption a no-brainer, when they work. Corporations stung by reports and by requirements from credit card issuers are also clearly protecting their networks better, although I'm sure we'll still see breaches at those firms that didn't cross every "t."
The 802.11n standard's emergence into an interim certified Wi-Fi state was also a significant milestone for faster wireless networking. Shipments of Draft 802.11n products in 2007 increased significantly, while prices dropped so much that it makes perfect sense to purchase a $50 to $80 Draft N router than a comparable G unit. Manufacturers made it clear as the year progressed that hardware sold today should generally be firmware upgradable to whatever the final, not much changed 802.11n standard is when approved in 2008.
Gadget-Fi continued on the rise, as an increasing array of devices included Wi-Fi as a connectivity option. Most notably, T-Mobile launched its HotSpot@Home service, the largest scale offering of converged cell/Wi-Fi calling. By year's end, they had four handsets for sale--two plain, a BlackBerry, and a clamshell--but subscriber numbers are unknown.
What's coming in 2008?
In-flight Internet (over Wi-Fi): 2008 is finally the year. It was supposed to be 2005. Or maybe 2002. But we should see a number of planes, mostly flying over the U.S., equipped with either in-flight Internet access or in-flight text messaging and text email. Connexion by Boeing's failure fortunately didn't discourage a half a dozen competitors who were in the R&D phase when Boeing wrote off its satellite-based Internet access venture.
AirCell, Row 44, OnAir, Aeromobile, Panasonic Avionics, and a T-Mobile consortium are among the announced or nearly announced firms with commitments or trials underway. AirCell and Row 44, focused on the U.S. market, plan to deliver Internet not voice to fuselages; OnAir and Aeromobile are working on mobile-based services, including voice, via existing cell phones and devices.
In 2008, American, Alaska, and Virgin America will launch trials over the U.S., and potentially move into production. OnAir should be expanding in Europe beyond the single French aircraft that's equipped in a trial now to RyanAir's fleet. And Aeromobile's Qantas trial could turn into real usage. There's likely action that will happen in Asia and the Middle East, too, that's not yet disclosed.
Other trends to watch
Wi-Fi in every smartphone with better integration. The iPhone was the leading edge, pun intended, offering 2.5G EDGE cell networking as part of the subscription price, along with seamless roaming to Wi-Fi networks. With RIM finally offering BlackBerry models with Wi-Fi, it's unlikely that any future smartphone model intended for serious users would lack the option.
Wi-Fi everywhere. Despite the setbacks in municipal Wi-Fi, wireless networks continue to expand, with better and better coverage found across larger areas and more locations. 2008 might be the year of hotspot saturation.
WiMax arrives. In 2008, we'll finally see production mobile WiMax in action in the U.S., and the questions about whether it works well enough and fast enough at the right price to beat current generation cell data networks, and make money for the disorganized Sprint Nextel will be answered. More certainly, Clearwire, with WiMax as its only option, will push aggressively to steal customers away from fixed, wired broadband, especially in markets with little competition.
Gadget-Fi a go-go. Wi-Fi will become an expected part of gaming consoles (already found in a few), cameras (found in crippled form in just a handful), regular cell phones (in dozens and dozens now), and music players (with more full functionality).
A pair of AP stories addresses Boeing's Connexion failure and in-flight etiquette for Internet access: The AP's Anick Jesdanun, who has written a series of detailed articles about the bloom in upcoming in-flight broadband and mobile services, files these two articles on specific aspects of the issue.
He writes about Connexion by Boeing, a service that more or less worked as advertised technologically--I heard many rumors about problems, but also knew many, many happy users--but that didn't have the right combination of weight and cost structure to "fly." As the article points out, and I've learned in the last year from extensive interviews, Boeing's system was out of date by the time it went up, and they had committed early on to extensive, expensive satellite contracts. Rivals and upstarts alike think they have the right combination.
Pricing is starting to be disclosed more and more. AirCell is clearly intending to charge about $10 for a domestic flight, which is pretty much what they estimated the cost would be all along. I expect there will be subsidies and deals for frequent fliers, potentially a monthly unlimited subscription (as AirCell has fixed bandwidth costs once the system is built), and partnerships with aggregators to lower costs for corporations.
In a related article, Jesdanun discusses whether and how airlines will deal with inappropriate behavior during conversations and in content viewing up in the sky. The various service providers will offer filtering of different kinds: Panasonic Avionics will filter for porn and violence, while AirCell will disable Internet telephony and voice chat. (Those who think you can get around that with a VPN or other purposes just need to remember that service providers can add jitter and such that will make calls indecipherable without affecting other sorts of data transmission.)
Broward County, Flor., consider county-wide Wi-Fi: The Ft. Lauderdale-containing county, with a population of 1.8m, is considering building a network. This cautionary article in the Miami Herald talks to a variety of local and national folks about the failure to anticipate complexities and costs in building such network. There is no county-wide network I know of that has proceeded very far due to service provider and utility pole problems. Miami-Dade County, for instance, had issues with gaining mounting rights and had to install new poles of their own--without disturbing sea turtle nesting sites.
Note the appearance of Verizon-dominated USIAA's David McClure who states, inaccurately, "Cities ran out literally by the hundreds to spend millions on engineering studies." Cities typically spent tens of thousands to hundreds of thousands for proposals; in the vast majority of cases, the service provider paid for engineering studies before building. McClure often has interesting insights, but should be disclosing his group's entanglement, which is used to mask bias towards large telecom interests.
Rural New Hampshire and Vermont (isn't it all?) looks into large-scale wireless: Wireless LINC of NH and VT is a plan to bring Internet access to the beautiful hinterland. The first stage of the network is planned at 400 sq mi in 2008. Rural areas have challenges (where to mount, where to get electricity, trees), but also the benefit of a lack of density, and often high points that are already wired with aircraft warning lights, cell towers, or other communications gear. They'll be testing different technologies and approaches.
The ultimate network size would be 6,000 sq mi and be run on a wholesale basis. The focus is on dial-up only users, rather than on competing with existing broadband service available in larger towns and cities. The project is estimated at $12m. That seems laughably small for such a large network, but given the particular focus and density, that might not be off base. It's also unclear how much equipment end users might require. A point-to-multipoint system would make much more sense in this environment.
Palm Beach Jewish temple has Wi-Fi: Temple Emanu-El has free Wi-Fi and encourages people to come, eat lunch, and work as part of a tool to foster community. No word as to whether the network has a timer that shuts it off from sundown Friday to sundown Saturday.
Spokane decides against expansion: The Eastern Washington city had an early large-scale hotzone courtesy of gear from Vivato. The network's size has remained at about 100 city blocks--ostensibly, no testing has been conducted I know of--and the city apparently has no real idea of the level of usage, to judge by remarks in the article. (Vivato's assets, by the way, seem to have be resold yet again, the article reports. I'm assuming there are patents that are of interest, as a $50 access point has many of the features only found in a multi-thousand-dollar Vivato unit a few years ago.)
Another silly Wi-Fi product from ThinkGeek: I love the retailer, but this item is just plain ridiculous. The Wi-Fi Detecting Watch has a display reminiscent of 20 years ago, coupled with a Wi-Fi detector. Not that it tells you the network name, cycles through what's available, or provides more details. Nope. Press a button and you get a number from 0 to 8 indicating signal strength. Signal strength of what? Hard to say.
The last of the first-generation analog cell networks are finally being powered down starting in Feb. 2008: Only about 1 percent of current cell phones (about 2.5m) are analog only. Automotive cell systems, however, were largely analog with model years 2003 and earlier; and home alarm systems with cell links apparently (says this AP story) didn't get digital radios until 2006!
Shuttle Express to launch Seattle-Portland van run with Wi-Fi onboard: Given the rural stretches along I-5, a route I know well, I'll be surprised if there's more than GPRS/1xRTT along a good hunk of the route. Still, at $180 round-trip for the 360 miles involved, it's barely more than gas and wear and tear on your car. Train service runs that route several times a day, but without an Internet hookup, and taking typically 5 to 6 hours instead of 3.
Minneapolis continues to shine through as perhaps the only big city Wi-Fi project that hasn't gone off the rails: The independent firm operating the network, US Internet, tells the Star Tribune that it's on track to what should be characterized as positive cash flow based on their description by spring. They have 5,000 customers now for their network which covers 30 percent of the city. But more impressively, 7,000 people have registered for the service when it becomes available. Their magic number for cash flow is 10,000 subscribers.
These goals are in line with signups seen elsewhere in smaller networks that often get worse reviews. The current 5,000 subscribers would represent less than 5 percent of the population (roughly) in the area covered. With 100 percent city coverage, 10,000 subscribers is just 2.6 percent of the nearly 400,000 population (2000 census). The network is now expected to be built out to its full extent by around February. (Update: To clarify, based on a comment emailed to me, that's 5,000 subscribers in 30 percent of the city, thus nearly 5 percent of that subset of population currently covered, not 5 percent of the 400,000 people.)
That's a far cry from the 15 to 35 percent uptake that some providers were predicting or requiring, although that was often in cities with much less broadband connectivity than Minneapolis's incumbents offer its residents. If US Internet can really hit a sustainable--not profitable, but sustainable--point with that few subscribers, they should be in a good position moving forward. The company also has city contracts, which is nearly unique in the metro-scale market, despite the current focus on obtaining "anchor tenancy" for new networks.
Reports indicate that after some tweaking a few months ago, the service performs quite well. Novarum has tested the network and hasn't released details yet, but says its impressions are favorable.
So what did US Internet do in Minneapolis that EarthLink and others failed to do elsewhere? Apparently, they planned the network with greater density, so that even while costs went up, they didn't double or triple. They secured a contract for services from the city. They obtained an advanced payment against services from the city. They responded quickly in an emergency--the bridge collapsed--earning enormous goodwill, while showing the flexibility of their system.
The only other comparable project is Portland, Ore., in terms of scale; that network reportedly has much more usage than any other metro network, mostly because it's an ad-supported, free service. However, the operator MetroFi and the City of Portland revealed weeks ago that the company wouldn't extend the network without an anchor tenancy commitment (Portland says no) or raising additional capital. MetroFi hasn't discussed revenue relative to expense or cash flow.
The first production in-flight mobile system is now airborne: After years of setting dates and pushing them back, OnAir has its first bird aloft. The satellite-based mobile phone and data service is installed in a test mode on an Airbus 318 (equipped for the purpose at the factory) owned by Air France. The plane was received in spring, but the necessary permissions took longer to secure.
The service is limited to text messaging and mobile email for the first half of the six-month trial. In three months, voice service will be enabled, but according to interviews I've conducted over the last couple of years, the voice part could be summarily disabled during flights or for the remainder of the trial depending on reception.
Calls may be made only above 10,000 feet (3,000m). OnAir said they have arrangements with several cell carriers--the company's service is like its own cell carrier in the air--including Orange, Bouyges Telecom, and SFR. The cost is "comparable" to international calling, and is estimated to be about US$2.50 per minute in previous reports and interviews.
I still recall a trip via high-speed rail I took in France in 2000 where, when a Francophone was talking (quietly) on his cell phone in the main first-class compartment, someone walked up to him and said, in French, why are you talking here? Why are you disturbing everyone? Do you think you're an American? The man protested feebly, and then got up to make his calls from the enclosed foyer of the car.
I have my doubts about whether in-flight calling could be truly acceptable outside of certain societies in which that kind of private space in public circumstances simply isn't recognized.
The Haj is unwired: The annual pilgrimage to sacred sites in Saudi Arabia now comes with free Wi-Fi. The service, paid for by the government and handled by local telecom services, is designed to help pilgrims better understand the Haj and connect with their muftis for counseling. The Haj takes place from Dec. 17 to Dec. 21 this year.
Interesting bidders in 700 MHz auction: The FCC released lists of accepted bidders and those who applications are incomplete for this coming January's showdown. The 700 MHz auction covers a set of coveted national licenses (six in total), that Google and others are bidding on, as well as a host of regional licenses and a combined public safety/commercial network that Frontline is a leading contender to win. You'll find Google, Verizon, and AT&T on the list, as well as the Paul Bunyon Telephone Cooperative. Some of the regional licenses are tiny, and could be huge boons for rural telcos, who, if their revenue is low enough, also receive bidding credits.
St. Louis Park, Minn., has moved forward on canceling its deal with Arinc for that firm to continue to unwire the city: In one of the few cases nationally in which a city was to own the network, St. Louis Park finds itself dissatisfied with the pace and quality of the network that Arinc is building. Arinc disagrees with the quality measurement, and argues that the pace was in part set by the city; it would now peg August 2008 instead of summer 2007 as the network's completion date. That is if the contract weren't being terminated by the city.
The Star Tribune reports that the city has found Arinc in default, and will proceed on the next steps to extricating themselves, which could involve payments and/or lawsuits.
One could argue that there's nothing here that reflects mistakes by the city based on an ownership model. Rather, their vendor, which isn't known for building Wi-Fi networks, wasn't reportedly able to reach the marks that the city wanted. The flexibility here is that the city can fire the vendor and find another. With the private ownership model characterized by EarthLink, cities have no recourse if the vendor pulls up stakes unless they have penalty clauses in the contract. Here, St. Louis Park is on the hook for lost revenue and the outlay, although they may recover some money from the vendor.
Watch for St. Louis Park to be included in future anti-municipal wireless reports, but also watch for any future changes to the way the network is built and run to not be included, no matter how positive.
Dash is accepting pre-orders for its $600 subscription-based navigation device with Internet connectivity: Using GPS for location and GPRS and Wi-Fi for connectivity, the Dash Express constantly updates traffic data from its own sources and other Dash devices--unclear on quite how. The Dash Express combines connected PDA features with mapping and navigation. The search is driven by Yahoo Local. You and others can email addresses straight to the device. Monthly fees are $10 with a two-year contract and $13 per month without one. The color screen is 480 by 272 pixels measuring 4.3 inches diagonally. Battery life is two hours but it comes with a car adapter, naturally.
Sprint picks new CEO: Wireline spinoff Embarq's CEO poached--after Embarq asked them not to, the Wall Street Journal reported a few weeks ago--to be the new CEO of Sprint. He had previously headed Sprint's local telecom division, so he understands wire in the ground. Will he get the right mojo going when no copper or fiber is involved? All eyes turn to Kansas. The future of WiMax at Sprint? It's hard to imagine it will be unchanged, given that CEOs have to put their stamp on affairs that are in disorder; they aren't hired to keep the status quo when sales are down and the future murky.
As the Denver airport goes free, its downtown experiment starts charging: The Denver Post reports that system operator Kiva Networks couldn't get the return it wanted from advertising-supported Wi-Fi supplemented by fees paid by the Downtown Denver Partnership. It's moving to a $5 per day, $25 per month rate, still collecting some subsidies from the business group.
New SOHO networking blog at ZDNet: Rik Fairlie has launched a blog at ZDNet specializing in small office/home office networking. He's got a strong reviews background, so look to the blog as a tool for more information about products that might help a network operator better and faster.
The portable Wi-Fi instant messaging device will offer cheap text messaging (SMS), too: The Zipit Wireless Messenger 2 or Z2 will gain a monthly service plan that allows 1,500 incoming and 1,500 outgoing messages for $4.95. The service will launch in a free trial period Dec. 20, and run through Jan. 31, 2008. The monthly fees would start on Feb. 1. Z2 works with what they describe as "more than 20 cellular carriers," so one assumes that's all the major carriers in the U.S. The device also interoperates with AOL, MSN, and Yahoo IM services.
The Z2 uses Wi-Fi for connectivity, and through a deal with Wayport, has free access at 9,000 McDonald's locations in the U.S. that Wayport operates Wi-Fi at. The Zipit Web site is essentially unusable for adults, designed to appeal to kids with a form of brain damage that doesn't allow them to navigate using menus or simple selection, apparently. It costs $150 and there's free overnight shipping through Dec. 21.
The single most sensible account of the state of Philadelphia's Wi-Fi system to date: The editorial board of the Philadelphia Inquirer popped this one out of the ballpark. I'm sure there are political wheels within wheels here--criticizing technically illiterate city councilman Frank Rizzo, praising Mayor Street's vision, cautioning the incoming mayor. But, holy smokes, this is a nice essay on why Philadelphia didn't wind up with egg on its face, or dollars out of its pockets.
The editorial notes the problems: the network's not done, it was late and over budget, it doesn't work well for everyone, and so on. But it reminds readers that 600 households have gotten computers and training, that EarthLink ate all the expense, and that the city's network forced lower broadband prices through competition.
The city of St. Louis Park, Minn., finds itself with a network that's incomplete and doesn't meet basic contract requirements, the Star Tribune reports: The for-fee network, being built by ARINC but owned by the city, hasn't fulfilled even the first level of service quality required before additional testing is performed in any of the four quadrants of the network, according to St. Louis Park. The network was supposed to be in full operation by summer, and now after lengthy delays, there's no longer any expected completion date. The firm was the low bidder in part because of its unique reliance on solar power to drive the nodes; that's apparently not the root of the problems.
The 45,000 person city has 20,000 households, and has 4,000 people reregistered for service, which likely means about one person per household. This is an incredibly high potential penetration rate, although the city's numbers at the time of the bid award were that they were looking for about 6,100 households (32 percent) to sign up within 12 months.
The city claims $300,000 in revenue loss so far due to the delays. Early in the year, residents balked at the first design for the 16-foot poles that would be mounted in hundreds of places around the city, and that led to a two-month hiatus while the infrastructure was redesigned.
EarthLink tells council via piece of paper it spent $20m on network to date: The entire network was once estimated to cost $20m to build, and somewhere around 75 percent has been activated, although (according to one correspondent who attended the meeting) what that 75 percent means is up in the air--75 percent of area or population?
If one recalls, EarthLink and its networking supplier Tropos were once talking about 20 to 25 Wi-Fi nodes per square mile. Tropos as one time charged thousands of dollars per node, although my sources indicate that the wholesale price, especially in volume, is dramatically below that. Even so, double network density increases the cost of finding real estate, rental costs for such, and has all the additional associated per-node and per-cluster costs.
The company didn't attend the meeting, sending an unsigned statement and leaving Greg Goldman, the head of the Wireless Philadelphia non-profit that supervises the network and the relationship with EarthLink as the target of councilmanic questioning. (His prepared testimony is online.)
Although EarthLink apparently wrote that the network is "substantially complete," that doesn't appear to jibe with the 75 percent number. Goldman said EarthLink had "several thousand" subscribers; WP has 613 participations in their program for free and discounted access.
WP has now secured $1m independent of the fees and expenses associated with overseeing the network for digital inclusion programs.
The Associated Press's Anick Jesdanun examines the fear of regaining in-flight connectivity among the jetset: This is a pretty hilarious, almost contrarian article about the reactions some always-on people have about gaining email, BlackBerry, and full Internet access while in flight. Some Type Triple A personalities look forward to more productive work time; others sigh about the last place away from the office getting a cord plugged in.
Public policy group releases report critical of Philadelphia Wi-Fi network, but not for reasons you think: Their conclusion is that the city-created non-profit Wireless Philadelphia should have built and owned the network. Long-time readers of this site will recall that I said many times when thinktanks and reports directly and indirectly funded by telecom incumbents made specious, unsupportable statements, that I would equally hold those in favor of municipally funded or authorized networks to the same standard? Okay, then.
My main take: The report looks to Philadelphia as a case study of what went wrong in municipal wireless, starting as the premise that public ownership with private operators working as contractors is what was called for by Phila. stakeholders and common sense, and that the non-profit that should have run the network has accomplished little. This view ignores the fiscal realities Phila. and most other cities face, the reasonable (as opposed to unreasonable) criticism raised at the time and that proved justified about putting taxpayer dollars into networks that weren't well understood how to build, and that the non-profit has, despite the long delays and now jeopardized future of the network, achieved some milestones through hard effort.
I'll also note that as with the 2005/2006 era anti-muni-wireless reports, the authors of this screed relied on secondary, tertiary, and even quaternary sources for their conclusions. Sources are often quoted or cited from newspaper accounts, or accounts of accounts in which a reporter or blogger (like yours truly) has summarized an opinion or quote read elsewhere. Trusting secondary or more distant sources mean leveraging your point on other people's conclusions or reporting skills, which is intellectually weak. (As a blogger, that's part of my trade, which I leaven with actual interviews, informal conversations, and constant email.) Speaking to everyone would have been overkill, but with some key players (or noting that they declined to be interviewed) would have provided the perspective that only human conversation introduces into ideology.
Now onto the details.
The thesis: Wireless Philadelphia essentially abrogated its role in supervising the network by handing over control to EarthLink. This doesn't conform to my reading of the documents or what's happened since. WP had no method by which to raise the money necessary to own and operate the network, nor to pay the fees required by other bidders.
WP is generally deprecated for failing to achieve goals that this report says were set for it. I haven't followed the digital divide side of things carefully, but I would argue that given three realities, WP had few choices: no money or near-term sources; an offer of a free network; a plan to move money from the no-cost network to WP. How could they turn this down, especially in 2005? The fact that little money flowed through the system doesn't appear to me to be the fault of WP or Philadelphia's leadership.
Page 3: "over-reliance on the unverified claims of companies attempting to run municipal networks, can lead to the failure of these initiatives": This is the worst form of couch quarterbacking I have ever seen. There were a lot of problems in how the business and technical models were constructed to run metro-scale networks, but this takes the cake. Of course the claims were unverified: No one had built networks of these scales before; WP would have been in the same position as EarthLink had it attempted to finance a network. Worse, even, because it would have run out of money long before the network had reached this point due to political expediency.
Page 4: "Other cities appear to have learned from Philadelphia's mistakes. Boston, for example, has embraced nonprofit ownership and a wholesale access model similar in key respects to what Philadelphia had originally envisioned." Yeah, and how has that worked for ya? Boston has so far shown no success in attracting any funding to what I agree was an interesting alternative to what was being proposed and accepted elsewhere. See more below.
Page 7: "Instead of a nonprofit owner leasing access to many companies, a single private corporation would own and operate the network, sharing a portion of its revenues with the nonprofit.": I would have urged the authors to have considered the notion that the fact that the former was infeasible may have led to choice of the latter.
Also, "Up to that point [EarthLink's bid]--and as the Philadelphia proposal originally outlined--it was assumed that a bond offering, private donations, or funding from the municipality would be required to pay for new construction." That's a bit of wishful retrospective thinking. A bond offering wasn't in the cards, even if it had been floated, due to the strong opposition of putting city credit in the mix; likewise, the municipality never had any interest in funding the venture, even if it became a customer of it. Donations, grants, and loans were deemed at one point to be the sole method of funding. (See the crossed-out part in my post from 03-Oct-05 in which I note EarthLink winning the Phila. bid.) Later, on page 14, the report states exactly what I note, which seems contradictory.
There follows a lengthy, detailed, seemingly accurate-in-the-facts (not the editorializing) history of the process of formulating the network's structure and bidding it out. It is troubling to learn how much (page 23) WP deferred payments on various expenses to have initial funding given EarthLink's potential withdrawal. I did not know (or perhaps just don't recall at this remove) that WP agreed to pay part of the electrical expense for EarthLink's network.
Page 24: The table of Municipal Wireless Models includes five entries that each have their own peculiarities. Minneapolis has been widely criticized by many authors cited within this report for not talking to stakeholders and bidding out a network with a much smaller process. The network builder has, however, made great progress, and network quality reports are quite good. It's the only true big-city success story that's almost unconditionally so for the current scope of build out. (Toronto would be another, but for a much smaller area with no firm expansion plans, despite statements by Toronto Hydro that they plan to expand.)
Boston, well, I've already explained that. Seattle is listed here even though as the explanatory text notes it's a fiber plan, not a wireless one. Wi-Fi was a buried option in the RFP, and after more than a year, there's been no public statements about progress with the nearly 20 respondents. It's gone nowhere. (I live in Seattle.)
Washington, D.C., may have an RFP, but I haven't heard about bidders. San Francisco is listed as "proposed," but it's an odd thing to show.
Was this table supposed to be "One Working Project and Four Stalled Ones?"
Page 25: There's a viable critique here about the cost of entry of any firm that wanted to purchase wholesale access from EarthLink. The bar to me was set too high, too.
Starting on page 27, the report starts to delve into WP's role and operations, which, as I've said before, I'm no expert on. WP released a statement prior to the report's release noting that, "None of the key players from Wireless Philadelphia - members of the Wireless Philadelphia Executive Committee, who were involved during the organization’s formation period on which the publication focuses; consultants; current staff and board; business partners; community partners; nor retail or digital inclusion customers--were contacted or interviewed by the author as the publication was being prepared."
This was a complaint I made about the early anti-muni reports: all the information was gathered from online sources or published reports, and not confirmed against interviews or factchecked with anyone involved previously or currently in any of the projects. Note my top-level comment about relying on non-primary sources, too.
Page 29: About EarthLink's late CEO Garry Betty: "In the summer of 2007, the CEO of the company died unexpectedly." That would be January 2007, not summer. Makes a great deal of difference in the context of EarthLink's decisions, especially noting that the new CEO, hired a few months after Betty's death, "immediately" changed the company's plans rather than months after that. Layoffs followed even later.
The report ends with conclusions I hardly find justified--conclusions that were made by drilling holes into pre-chosen points in the narrative and then inserting pegs in those holes to hang the conclusions. There's a lot of logical leaps made here, with the same lack of rigor I've criticized in the no-public-ownership crowd. I have often suggested that these kinds of reports could be better received if they weren't so ideologically rigid: the inclusion of facts that do not support one's precise position represent a robust point of view; the exclusion, a timid one.
The appendixes show more of this. In a discussion of Boston's unique model, which I've noted is troubled, the report notes on page 80 that, "As of early June 2007, a number of companies were competing to play a role in the pilot Wi-Fi project in the square-mile Grove Hall neighborhood. Boston is on track for citywide Wi-Fi by the end of 2008, Pamela Reeve the chief executive of OpenAirBoston, the nonprofit created by the city to manage the wireless project, said at the time."
They footnote this to an article in the Boston Globe, which, in its second paragraph, states, "But the ambitious plan to create an "open access" wireless Internet environment citywide has hit some speed bumps, including the absence of major donors, delays in launching the pilot project, and blogger protests about Web filtering at WiFi hot spots." And that Reeve, after saying the project is on track for 2008, also agreed, "the target date is 'aggressive'" as well as admitting "Boston has yet to line up the 'key bucks' funders, such as companies, universities, or hospitals, it is hoping for."
A more honest assessment of Boston would have included those negatives.
Page 36–37 cover Corpus Christi, a city that has its Wi-Fi network now operated by EarthLink, and which carries no mention of EarthLink's statements about the future of networks in cities it's already operating in, which I find a strange admission, as if this appendix was written months earlier and not revisited.
Pages 40–41 discuss the Wireless Silicon Valley project, but only notes at the end that there is a coordination issue with Google's network in Mountain View, rather than the fact that the project has made no progress in some time, and Azulstar was unable to raise funds. (It was pulled as the lead on the project recently, but I can't critique the report for not including that particular fact.)
I was not asked to review this prior to publication, and had I, I would have just as forcibly presented these objections and more.
With the network's future in flux, EarthLink sends unspecific note, not representative, to council meeting: Philadelphia city council members were a bit irritated, in this AP report on the hearing, to not get any specifics or a warm body from EarthLink about the future of the network, or even present details. The AP reports that EarthLink's note indicates it's talking to providers who might "have an interest in this network."
Over the last few months, EarthLink has backed off, canceled, and wound down most of its municipal commitments. The networks it's actually operating or was in the process of building are likely to be shuttered or sold unless accommodation can be reached with cities to purchase services beyond that agreed to when contracts were signed.
Hey, we finally got numbers about Tempe, Ariz., in an article about the network operator's assets being purchased: I'm not sure what Telscape (incorrectly noted as Telescape in the article) Communications is thinking about. The coverage of the current network after three years of work is apparently inadequate, despite what is now called Kite Networks' increasing from 400 to 1,200 nodes to cover the city's 40 sq mi. (They don't say 1,200, but they said 400 nodes originally at launch, 600 later after coverage was too poor, and "three times" the original count in this article.)
Telscape is a bilingual telephone company and has operations in parts of the country with large Hispanic audiences. Thus, Tempe isn't a bad place for them to concentrate on. The firm's chairman told Tempe's city council he was considering a buyout of the network because he thought he could market the network much more effectively, obtaining 10,000 customers, which would be 6 percent of Tempe's population. Seems unlikely. Wi-Fi market penetration hasn't exceeded low single digits yet in any city offering it unless the service is free; and free doesn't stop people complaining about it. In Tempe, particularly, the university population is concentrated, but fleeting and cheap (students don't have an extra $10 or $20 per month), while served in part by Internet access in dorms and on campus. (The university, last time I checked, has no campus-wide Wi-Fi.)
The company is researching the deal, but the chair thought the network was well built technologically. We'll see what due diligence brings. Meanwhile, recall that Kite Networks was bought by Gobility in a deal that required additional funds to be raised that weren't. So I'm not sure in the end (there weren't any filings recently by Kite's former parent company) whether Gobility will be selling the assets or returning them to NeoReach for them to be sold.
I wrote about Tempe back in March 2006 for a feature in The Economist in which I discussed all the things that would be problematic for Wi-Fi networks and that weren't being addressed. I noted that Tempe already had a wireless network used for backhaul that had dropped their telecom costs from $1.7m to $0.5m a year (exclusive of capital). Tempe was supposed to be an ideal case, given its compact audience and rather flat terrain, as well as the large university population.
In a bit of understatement, here's my conclusion from the Economist article: "The real measure of municipal wireless networks will not be in places such as Tempe, where they are expected to work, but in bigger cities, where success is far from certain. Whether service providers will be able to meet the required technical standards and still make a profit will soon become clear."
Washington's capital, Olympia, opts out of Wi-Fi: They decided the small sum they'd allotted was too small to produce any interesting results. They had budgeted $20,000 for some amount of Wi-Fi, and learned it would cost $100,000s to build what they'd envisioned.
Philomath will get Wi-Fi, perhaps for free: A local provider might get a contract from the city which would enable it to build a network that could perhaps be offered at no cost. The same firm built a free, small hotzone in Corvallis, where the state's agricultural and engineering college, Oregon State University, is based, and they may be able to expand that network. The small Oregon city is pronounced fill-OH-muth, not, as you probably just said to yourself, FILL-a-math. (I spent formative years in Oregon.)
JetBlue launches free email, IM on Dec. 11 with Yahoo, RIM; no Internet access or attachments: JetBlue purchased a slender 1 MHz of spectrum in an FCC auction last June, and they plan to use it to provide their Wi-Fi laptop and handheld equipped passengers with free instant messaging and email service in partnership with Yahoo. Attachments won't be allowed. Owners of two models of BlackBerry with Wi-Fi will be able to use those for email as well, and one hopes that the security options that have made Research in Motion popular will be preserved. (You can read a summary of the auction in a post I wrote in June 2006 at the conclusion of bidding.)
JetBlue will launch the service using an Airbus A320 they named BetaBlue as it's also used to test entertainment offerings by its LiveTV subsidiary, which was technically the firm that bid on spectrum.
While AirCell was the bigger winner, buying 3 MHz over which they plan to launch broadband service (over 1 Mbps in each direction) with American Airlines and Virgin America next year, JetBlue's 1 MHz is nothing to sneeze at. It just can't handle Web surfing. An onboard media server with cached sites and streaming video could be a nice complement.
Update: The New York Times has some additional information, with the reporter having flown on a test flight. She notes that the service had handoff problems among ground stations (which she calls cell sites; not technically incorrect, but less descriptive), and that the handoff factor was one reason JetBlue wasn't going to charge initially. (Those with long memories will remember JetDirect, a short-lived service on airlines like United that used Tenzing's system to proxy mail through an onboard service, using AirFone's narrowband service for backhaul. The proxy allowed discontinuity in service. Tenzing merged into OnAir, which has focused on satellite access, and may launch its first equipped aircraft by spring 2008 after nearly three years of delays from their original plans.)
It's worth noting that the restriction in using Yahoo's mail service means that corporate networks that require virtual private network (VPN) connections to gain access won't be supported, and companies and agencies that require secure email are out of luck. This systems includes no cell aspects, so handhelds and smartphones without Wi-Fi have no chance of working.
Flight 641 leaves JFK on Tuesday at 8 am, but that's the only scheduled flight for that particular aircraft. Depending on feedback, JetBlue "hopes" to install the system on its whole fleet.
The House overwhelmingly passed a bill that puts a huge onus on individuals and ISPs to report a broad range of images or face $150,000 or $300,000 fines: The bill, if passed, would require anyone offering network access (under a broad definition) to report any images they come across that covers a broad range of potential depictions of abuse of minors or their exploitation. Declan McCullagh notes in his post at Cnet that it's so broad as to include photographs, drawings, and cartoons that require interpretation as to whether they would meet the test. Sounds like viewing an Abercrombie and Fitch catalog could qualify.
From McCullagh's report, it's clear that if you operate, for instance, an open Wi-Fi network from your home, you're not obliged to monitor other people's uses of it. However, if you were in any fashion to become aware of behavior circumscribed in this law, you could be fined $150,000 on a first offense and $300,000 on a subsequent one--unless you preemptively report, in which case you're immune to lawsuits and prosecution. Which means people are going to report a lot more or shut down otherwise free hotspots.
Reporting requires that you provide the images that you saw, which obviously opens people up to charges themselves unless everything is handled extremely carefully as possession of child pornography is a de facto crime that allows for no explanation in much of the U.S.
The most typical scenario as I can see it is a single-location coffeeshop that provides free Wi-Fi. A barista is walking through the shop and catches a glimpse of an image they think is problematic before the user shuts their laptop or navigates to another page. They aren't sure, so they tell the manager. The manager then, if he or she doesn't immediately report and attempt to ascertain as much information as possible, could face the fine (or perhaps the store's owner), if I suppose the FBI or a local crime unit tracks back a person's usage to the store. But how to track that usage back and prove someone saw something and didn't report it?
This is all sort of a pisser, to be frank, because there's practically nothing more appalling than child abuse and child pornography. And, thus, I hate to be in a position to explain why a law to protect children erodes individuals' sense of security in not facing massive fines or arrest when they may not be in a position to act precisely how a law demands. Surely, people who see bad stuff should report it, and many do.
The bill, as written, was rushed into session using an expedited procedure that avoided the usual review. Declan McCullagh writes at Cnet that the legislation also changed substantially before being brought to a vote. So reasonable opposition didn't form before its passage. The Senate could approve a similar view, but it's hard to know how enforceable it will be.
IDG News Service reports that Azulstar was unable to raise the funds necessary to build the network, but will continue as applications provider: Big-name partners IBM and Cisco are seeking a replacement firm, which seems essentially a doomed effort. Who would step in at this point on a project that has no committed municipal partners, no test networks build, a failed effort to raise funds, and no input into how the project was shaped? That would be...a chump.
What should happen is that this attempt should be scrapped, and one city should volunteer to work with the group with pilot funding chipped in by counties, Cisco, and IBM, as well as other cities, to become the model testbed for a network with specific goals set for specific services. But that's not going to happen. There's too much committed to this effort to not see it through--and to fail spectacularly before a re-think happens.
The Houston Chronicle notes that the city isn't 100-percent sure that their Wi-Fi network won't be built: In the meantime, they're using the $5m penalty paid by EarthLink for a late start on the network to build neighborhood hotspots, and pay for training and hardware for residents ($3.5m), and to build out public safety and other wireless project ($1.5m).
One of the first places you could obtain Wi-Fi access in airports was in American Airlines's Admirals Club; now, it's free: The airline is still working with T-Mobile, but is providing this amenity at no cost to members of its Admirals [sic] Club and one-day pass guests. The frequent flyer membership number (for club members) or a unique PIN (for day pass guests) gets you in. Given that T-Mobile charges $20 to $40 per month for its HotSpot subscription or $10 for a day, the $300 to $450 annual club membership cost doesn't sound like a bad price if you don't otherwise need T-Mobile service or rely on other offerings.
NS, the Dutch train operator, suspends trials of onboard Internet, cancels plans (press release in Dutch): The operator, which was considering putting Internet service on hundreds of trains, says that the potential advertising revenue that they had planned to use to cover the service isn't enough, and that the costs of installing service exceeded their expectations. They're shutting down a trial of Internet access and train information that they've had underway for some months. They're refocused just on train information (like arrival times), which has smaller overhead.
This is yet another indication of the difficulty of providing Internet access to trains, but not the impossibility of it. Rail lines can opt for line-of-sight WiMax or equivalents, cell data networks, or satellite for backhaul. No option is cheap or easy, but all are possibilities.
Back in September, I noted that the three train lines in central California considering Internet service or having once deployed it--ACE, Caltrain, and Capitol Corridor--are currently all in abatement. ACE still has a coming soon banner on their site after many months; Caltrain decided the bids for service weren't a good deal for them; and Capitol Corridor never launched its pilot tests due to problems with bidders. Update: I was too hasty on Capitol Corridor's plans. They were able to participate in a joint test with Caltrain which, although costs weren't to Caltrain's liking, could work for the CC Joint Powers Authority that runs the rail. The agency documented its plan, which involves working closely with Union Pacific Railroad, for a full deployment by 2009 if all goes well in negotiations and planning.
When I wrote about train-Fi for the Economist on 21 Sept 2006, I had spoken and corresponded with a good 20 people in all aspects of the industry--service providers, private operators, train authorities, consultants, and so on. The general buzz then was that with a number of service providers working hard, that perhaps several hundred trains would be equipped with service by now. We're probably still at just above the same number as last year: perhaps 100 trains (not cars, but operating engines on lines) worldwide having real Internet service.
A catalyst might be needed, like a new, less expensive broadband satellite offering, cell operators extending their service specifically for trains along certain corridors, or simply more work on WiMax and other licensed technology that could be deployed in the right places.
Boingo wireless offers GoBoingo!, a 1 MB application for connecting to its aggregated network of over 100,000 hotspots: Glenn, you might say to me, even though we might not be on a first-name basis, why in the name of all that is right are you posting an entire item about Boingo Wireless's new connection software for Windows 2000, XP, and Vista? Like we care, you may say.
Here's the deal: Boingo isn't precisely poised for world domination, but they're in a very interesting position. They're the only company in the world with both a worldwide hotspot footprint that's quite enormous, and that charges a flat rate for both U.S. only and worldwide access. Thus, this move is in line with their efforts to make it easier to get folks on the go--Go!--in the many airports they service to download a client quickly, install it without a reboot, and gain access to the network.
Boingo has leveraged its airport network into a significant but undisclosed increased number of subscribers. The easier it is for a traveler to get special software installed that lets them become a trial Boingo user, the more likely that user converts. Boingo has offered a similar package for Mac users--still not updated for Intel Core 2 processors, for shame--and they will continue to offer a full connection management package for Windows that includes network key management and an 802.1X supplicant.
With the proliferation of handheld devices with Wi-Fi built in, most of which lack a browser or are intended for an audience that shouldn't have to mess with logging into a hotspot--think about the Kindle, where Amazon chose cell data partly for simplicity--Boingo has a big opening into a big market.
So you might care, just a little bit.