Kachingle is a way for regular readers to support this site, as well as others they visit frequently.
Receive new posts as email.
| Sun | Mon | Tues | Wed | Thurs | Fri | Sat |
|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | 6 | |
| 7 | 8 | 9 | 10 | 11 | 12 | 13 |
| 14 | 15 | 16 | 17 | 18 | 19 | 20 |
| 21 | 22 | 23 | 24 | 25 | 26 | 27 |
| 28 | 29 | 30 | 31 |
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.
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.
Powered by
Movable Type
I somehow missed a significant change in Starbucks revised rewards program launched last December: In the initial announcements from Starbucks about a reworking of its discount card (eliminated) and rewards attached to its stored-value Starbucks Card (revamped into tiers), I reported on 3 November 2009 that new card users would have to make five purchases to reach a new "Green" level, and be able to use two continuous hours of Wi-Fi each day at no cost, along with a bunch of free beverage add-ons. I also wrote that Wi-Fi access was enabled only for 30 days following the use of card. Neither turned out to be true in the program that launched.
But when I checked a couple of days ago for a forthcoming article in The Seattle Times, I found different verbiage--in fact, two different sets. On the main rewards page, Starbucks said the deal was as above; on the FAQ, the Wi-Fi deal was much simpler. After a couple of days of finding the right person at Starbucks, I've got the answer: the program changed, and they didn't make a fuss about it.
The deal as it stands requires just obtaining a Starbucks Card and making a single purchase or adding $5 to the card's balance, which qualifies you for the "Welcome" level. You must then register the card, and sign up for an AT&T account. After that, Wi-Fi is free for two successive hours a day. No further purchases are needed. The Web site is being fixed (you're welcome).
That's much closer to free than Starbucks has offered before, and a big turnaround from what it was promoting in its FAQ in November.
It's nearly free, even.
A huge swath of AT&T customers as well as Qwest DSL customers get access to Starbucks all day, every day, at no additional cost.
Posted by Glenn Fleishman at 2:08 PM | Permanent Link | Categories: Free, Hot Spot | No Comments
Skype is promoting its Skype Access pay-per-minute Wi-Fi option by making it free this weekend: From the start 20 March (0000 GMT) to the end of 21 March (2359 GMT), you can use Skype Access to access 100,000 hotspots worldwide at no cost. The service, available since last July, must just have left beta testing, even though it was unclear to me that it wasn't a released service.
Skype Access uses Skype software to handle a Wi-Fi login transaction, and normally costs (including Luxembourg's VAT) US$0.22, Cdn$0.26, €0.16 per minute; fees are debited from the Skype Credit in your account. Skype itself is a free download and free to use for Skype-to-Skype IM, file transfer, audio chat, and video chat.
That's expensive in the U.S., where you can pay $10/mo for unlimited Boingo Wireless service at the same or more locations, and where $8 to $12 is the most you'd pay for 24 hours (the equivalent of about 30 to 60 minutes at Skype's per-minute rate).
However, elsewhere in the world, being able to get 10 minutes of Wi-Fi for a couple bucks might seem much more appealing without having a service plan or paying ludicrous European hotel prices, which can be $20 to $40 per night.
A full list of supported hotspots networks is available.
Posted by Glenn Fleishman at 10:09 AM | Permanent Link | Categories: Free, Hot Spot, International, Roaming | No Comments
Here's today round-up of brief Wi-Fi items.
David Strom highlights the risks of in-flight Wi-Fi: David, who I have known for many years, writes about how much data people may wind up exposing on in-flight Wi-Fi networks, but not over the wireless network. It's a fascinating point of view. His seatmate on a recent flight revealed a ton of information to David's casual visual inspection, including passwords. He recommends privacy filters. 3M makes a host of these.
The Boston Globe looks at why Salem's free downtown Wi-Fi effort faltered: It's an interesting roam around that city, but doesn't precisely answer the question of why downtown businesses didn't continue to fund the group effort. Various shops have free Wi-Fi, but perhaps each wanted to have people come closer, instead of creating a commons. Boston's own OpenAirBoston municipal effort isn't even mentioned; the Boston Globe ran a long feature last August on the effort's small but interesting progress. (You can read the background about OpenAirBoston in this long article I wrote in 2006.)
Speaking of blasts from the past, the Washtenaw County effort appears dead: In Michigan, in the county that contains Ann Arbor, a long-running nearly unfunded private/public partnership has bit the dust, the local paper reports. The original firm, 20/20 Communications, bid on a plan with no funding to build it out, and a federal request for stimulus money was turned down. 20/20 apparently has sold the small number of current customers to 123Net, 20/20's president has left the firm, and the new company has no interest in county-wide service provision.
Posted by Glenn Fleishman at 11:30 AM | Permanent Link | Categories: Wee-Fi | No Comments | No TrackBacks
I'm not sure the Metropolitan Transportation Authority (MTA) has any chance of practical bids for their RFP: A long-awaited RFP is out for providing Wi-Fi service on all Metro-North and Long Island Railroad (LIRR) trains and most stations. Those rail lines accommodate nearly 600,000 passenger trips each week. (The PDF of the RFP has disappeared from the MTA site, but Google has a text cache.)
The MTA wants a service provider who would operate a network to bear all the expense of installation and operation (including railroad labor costs for same), provide 24x7 customer support, and uninterrupted service.
But the proposal is pretty muddled. While digital advertising (changeable signs on board trains and at stations) should be part of a bidder's thinking to minimize the cost in installing such systems, there's no spec for those systems. A bidder can build a bid partly around offering such services. The MTA also likes bids in which the authority shares in revenue.
I don't see how this could fly. No sensible firm would propose taking on all this expense without any assurance of revenue beyond the public Wi-Fi side of the system. Despite the large number of passengers, many of those most likely to pay already have 3G service on smartphones or through laptop cards. There's no operational services component, and that should be the baseline for any new rail RFP of the last five years.
It's not so much that 3G service works perfectly along the various part of the system, but it certainly works well enough. A service provider would either need to be a cell operator that can use the system to promote and sell Wi-Fi by itself and a combination of 3G and Wi-Fi (AT&T and T-Mobile notably in this position), or build on another technology that would go well together to feed service to trains and mobile devices (Clearwire's WiMax).
The system described would likely cost many tens of millions of dollars to build to the specifications that the MTA is requiring, without any substantial potential to reclaim that as revenue.
Posted by Glenn Fleishman at 9:03 PM | Permanent Link | Categories: Rails | No Comments | No TrackBacks
The firm that promised mobile everything for a low, low price but owned no network has disappeared: My friend Nancy Gohring at IDG News Service wrote a series of articles in mid-2009 about Zer01, a firm that said it was not a mobile virtual network operator, but somehow had access to a national network on which it would offer unlimited calling, mobile broadband, texting, and other features at a rate far below what operators charge. Unlimited mobile data seemed particularly impossible, given carriers cap at 5 GB for laptop use, and only a handful have specific unlimited smartphone (no tethering) data plans.
Nancy writes today that Zer01's Web site has gone dark, referring users to Google; its press spokesperson, Ron Dresner, didn't return calls and his Web site no longer lists Zer01 as a client; and Zer01 doesn't appear to be involved in the upcoming CTIA trade show.
I reiterate to anyone who doesn't know but will listen: all these deals that seem too good to be true are invariably too good to be true.
Meanwhile, the mainstream carriers now offer unlimited calling and texting plans that, for heavy users, are relatively inexpensive compared to previous plans that used pools of minutes and messages.
Posted by Glenn Fleishman at 3:03 PM | Permanent Link | Categories: Cluelessness | 2 Comments
Apple started taking pre-orders for the iPad this morning, and provided more details of the optional 3G data plan: Apple's deal with AT&T, assumed to be somewhat similar with non-US carriers when those plans are announced, called for $15 price tag on 250 MB of combined upstream and downstream data usage, and $30 per month for unlimited usage. Neither plan required any contract commitment.
[Update: To expand on a question and my answer in the comments, Apple and AT&T market the iPhone and iPad 3G unlimited plans as truly unlimited. Average iPhone user is a few hundred MBs per month; heaviest use is likely on Wi-Fi networks. With no tethering, an unlimited plan isn't a huge risk for a carrier. If AT&T ever lives up to its promise to offer tethering, expect a separate fee and 5 GB limit.]
Apple made clear today that the 250 MB plan doesn't tack on overage fees when you exceed that quantity of data: rather, you can either upgrade to unlimited (presumably for $15, but that's not stated), or shut off 3G when you run out. That's the most humane offer I've seen to date.
As stated at the iPad launch announcement, iPad owners with 3G built in ($130 more than the Wi-Fi-only version) can sign up for a plan or turn it off from the iPad without having to visit a Web site or go through a separate process. Reducing friction always improves sales, and that also dramatically reduces AT&T's costs by making it a self-service, Apple-handled option.
All four major US carriers offer a low-bandwidth option for 3G service in which 200 to 300 MB of usage is included, but extra megabytes are charged at 10 to 20 cents a piece--$100 to $200 per GB. Virgin Mobile is the only firm that lets you buy preset chunks of data (which must be used within either 10 days for the smallest increment or 30 days for the three larger options).
Apple says that 250 MB subscribers will be warned when they have 20, 10, and no data left, and can then choose to upgrade. Because the plans are month-to-month, a subscriber could upgrade one month and return to the 250 MB level the next.
The lack of "gotchas" will definitely go a long way in getting more people to buy the 3G model and use the service.
Posted by Glenn Fleishman at 11:20 AM | Permanent Link | Categories: 2.5G and 3G, Consumer Electronics | 2 Comments
One of the legacy muni-Fi networks will have new (or no) owners: Esme Vos writes at MuniWireless.com about the current state of the Riverside, Calif., network operated by AT&T. The network was the first and only bid by AT&T with MetroFi, which was unable to complete that network along with many others, and which shut down in 2008. In Riverside, AT&T kept up much of its end of the bargain, hiring Nokia Siemens to complete the network, which Vos says only reached 77 percent of the city. (One expects there's no SkyPilot gear left in place, either, but I don't know that for sure.)
The network has 20,000 daily users out of a population of about 300,000 (in 2000); the county has over 2.1 million residents.
AT&T wants to give the city the network at no cost, but the city is facing revenue shortfalls like the rest of the country (and most of the world). It's trying to get a federal grant.
Of the networks originally built in part or whole by EarthLink, Kite, and MetroFi, only a handful remain in operation. Philadelphia recently moved to take over the remains of the network there from an interim firm that had been planning to build out a variety of access services.
Posted by Glenn Fleishman at 10:50 AM | Permanent Link | Categories: Metro-Scale Networks, Municipal | No Comments
T-Mobile cuts its 3G prices by $10 per month for unsubsidized hardware: It's long been an irritation of cellular customers that even when they pay the full price of a phone or 3G modem, the monthly data charge remains the same as people who get a subsidized version. That's changing, with T-Mobile in the lead. T-Mobile cut its unsubsidized monthly fee for 3G service by $10 per month for its 200 MB and 5 GB plans.
Last month, T-Mobile dropped the monthly data cost for some smarpthones by $20 per month for people who purchase outright. The New York Times noted that the MyTouch 3G is $400 without subsidy or $150 with a two-year contact. The most basic voice and data plan is $60 per month for the unsubsidized phone or $80 per month with a subsidy--a $230 difference over two years.
The latest deal comes with the announcement of a ship date and price for the webConnect Rocket, an HSPA+ modem that works at a raw data rate of 21 Mbps. The new 3G service plans for 200 MB and 5 GB per month (combined upstream and downstream usage) will be $19.99 and $49.99 with a modem purchased at list price; the $29.99 and $59.99 remain in place for subsidized modems. The lower price also doesn't require a contract commitment.
T-Mobile sells its HSPA 7.2 webConnect modem for $129.99 retail and $19.99 with subsidy; that's a $110 difference upfront for $240 worth of savings over two years, plus the flexibility to cancel at any time.
Of course, T-Mobile also charges the highest rate for data overages: 20 cents per MB ($200 per GB) above the 200 MB or 5 GB limits of the plan. Virgin Mobile Broadband's prepay plan is $60 for 5 GB (used within 30 days), but an additional 5 GB is another $60 even within 30 days, while T-Mobile would add $1,000 to your bill for that privilege.
If T-Mobile wants to be truly progressive, it needs to sort out that discrepancy, designed to keep down usage rather than truly penalize subscribers. With HSPA+ offering potentially three times or greater the speed of the current network, limiting users to 5 GB and charging $20 per GB for overage seems very out of sync.
The Rocket, by the way, will cost 99.99 with a two-year contract; the outright purchase price wasn't announced.
Of course, the carriers have risks in carrying your subsidy over two years, plus early cancellations, the costs of churn, and so forth. By having you pay upfront, they don't carry the cost on their books, and it likely on average costs them less to handle you as a customer.
Posted by Glenn Fleishman at 10:35 AM | Permanent Link | Categories: 2.5G and 3G | No Comments
Amtrak has a request for qualifications to build a system-wide communications platform operational and passenger purposes (link not yet available): The RFQ looks to find vendors who could build a system for giving Wi-Fi to passengers, and running a host of operational data needs for the railroad.
The RFQ spells out platform requirements, which include inter-car communication via wireless signals--no wires connecting cars--and dynamically assembly of a network when cars change in a train. The system has to support the ability to use multiple cellular networks, aggregating in additional bandwidth as available (such as "external Wi-Fi and track side wireless networks"). Heterogeneity seems to be the message here: no single-platform/single-technology commitments, and a plan for simple module-swapping 4G migration must be included.
The RFQ asks for a number of specifics from vendors who choose to bid. I don't see any schedule information in the document I was provided.
The document gives some marvelous numbers: Amtrak carries 66,000 passengers a day; 38 percent travel for business, and 14 percent commute. Atlanta's airport, the busiest in the world, had 222,000 "emplanements" (counting stepping off and on separately) per day in 2009.
Posted by Glenn Fleishman at 12:31 PM | Permanent Link | Categories: Rails | No Comments
It is my pleasure to link to the finest mainstream article I've read on the quandary of whether there's a health risk from EMF radiation: I salute James Geary for not dismissing the concerns of people who are obviously suffering from something, for not pandering to those people, for not citing junk science, for not posing the issue as a "debate" between two sets of equally valid information, and for not ignoring all the uncomfortable issues around the edges that have not been fully explained.
It's a good read, partly for the people involved, and partly for the route Geary picks through the minefield to present good information to a mass audience.
I have two quite minor quibbles with the article. First, there have been dozens of studies on electrosensitivity, and all but a handful (which haven't be reproduced) show that self-identified sensitives cannot determine whether a signal is present or not. The article mentions this in passing, but the scope of work in this field is quite large. Second, the Interphone study as a whole is yet to be released, but multi-country components are out, and they generally confirm a lack of correlation between cancer and usage, with some exceptions that may get further study.
(Disclosure: I write for Popular Science on occasion, but I had nothing to do with this article.)
Posted by Glenn Fleishman at 9:29 AM | Permanent Link | Categories: Health | No Comments
I am compelled to write this story simply to say it does not matter: Reports came out a few days ago that all the iPhone OS applications that sniff out Wi-Fi, scanning the vicinity for signals and other information, have been removed from the App Store, the only authorized place from which iPhone and iPod touch owners can download apps, free or fee.
It doesn't matter, despite all the yelling about it. The sniffers were dropped because they use a private framework, hooks in the operating system that are not documented nor allowed for third-party developers to use. Apple scans and checks for these kinds of uses, and rejects programs that employ them. The sniffers got a pass for some reason, but someone at Apple woke up and kicked them out. It's a shame for the developers who put time into them, but using private frameworks is a completely well-known risk.
This dumping of sniffing apps is entirely distinct from Apple's arbitrary and capricious acts related to other programs and categories of programs, in which developers acting in good faith and according to guidelines find themselves on the wrong side of a shifting line. That happened to "sexy" programs, all of which not made by major firms like Playboy and Sports Illustrated, were dropped without warning.
It's been suggested that Apple should have an open and closed mode on the iPhone, letting people choose to run apps that haven't been reviewed and filtered by the company, but making no guarantees about those; in the closed mode, only Apple-approved apps would run. Apple seems to have no motivation to make that change, however, with its closed system working just fine for it, if not developers.
Posted by Glenn Fleishman at 1:02 PM | Permanent Link | Categories: Industry | 4 Comments
Loyal readers, I'm trying out a new way that you can support this site directly: Kachingle just launched, and it's the latest, but most interesting in my view, in a long series of ways in which individuals can push small amounts of money that aggregate into potentially large quantities without much effort. I've tried many of these over the years, but they typically involve too much work on the part of you, the reader.
The idea behind Kachingle can be explained in two sentences. You pay Kachingle $5 per month and choose which sites, when you visit them, that you want to support. Kachingle tracks your visits to those sites, and then divvies up your $5 proportionately among your supported sites based on your visits.
From the user standpoint, you get transparency. You can see how much money I make overall, and how your dollars and cents are being divided. At launch, Kachingle gets 10 percent and PayPal gets roughly 10 percent. As volume increases and other factors come into play, some of those fees will drop. Those fees are taken out after you pay, netting me 80 percent of whatever proportion I get.
This site is obviously a labor of love, but I will guarantee that revenue from advertising and other methods is directly proportionate to the amount of time I can afford to write original reporting and analysis. I'm trying Kachingle as one experiment to see whether individual readers of the site who find it useful can, through very small increments, boost revenue enough that I can devote more time to it.
Kachingle is in the early days, so there are few user and few sites using the service yet. That will change, clearly. And the experiment is limited to the $5 you spend each month if you sign up, which make the risk small.
To sign up for Kachingle use the badge at the upper left below the site logo, or follow the link in this post.
A nice side benefit of Kachingle is that the more people and sites that use the service, the more all sites that use it benefit. We can rise the tide for all boats.
Posted by Glenn Fleishman at 4:43 PM | Permanent Link | Categories: Administrative Detail | No Comments
Credant, a UK firm that sells data encryption tools, claims thieves sniff Wi-Fi in laptops stored in cars: I've been letting this percolate for a couple of days in my head, and would appreciate comments from those of you who know the nitty-gritty. Credant is claiming that thieves can use Wi-Fi detectors to find laptops in cars that have Wi-Fi active, because some laptops don't go to sleep for 30 minutes after the lid is closed or sleep is activated. (Thus, Credant says you need to have encryption software installed to prevent access to data, rather than, say, fix your system or add a car alarm.)
[Update: Eric Lai has a terrifically detailed article at Computerworld that addresses many of the questions below.]
Here's my problems with this scare-via-press release:
Any other ideas? Or is this just plain scaremongering?
Posted by Glenn Fleishman at 10:46 AM | Permanent Link | Categories: Security | 4 Comments
A couple of reports on the new service Amtrak launched in the Northeast: Amtrak's offering free Wi-Fi in six stations and on the Acela line that runs between Boston and D.C.
From the Washington Post, Rob Pegoraro runs through some of the details he's found out. On trains, streaming services are blocked, and some content is filtered, without complete disclosure. There's no excuse for avoiding full disclosure. Pegoraro saw rates of 1 Mbps down and 200 Kbps up from the aggregated mobile broadband service.
Nomad Digital's backend is being used, which can take signals from multiple 2G/3G operators to piece together continuous coverage. I imagine the firm uses a virtual network that uses proxies on both ends to allow a continuous IP connection regardless of the intervening network pieces. The user has no awareness of this, and remote sites maintain connections via the proxies.
The service in stations is quite a bit higher, with Pegoraro measuring 3 Mbps/600 Kbps. Regular correspondent Klaus Ernst, an inveterate tester of new Wi-Fi systems around Manhattan, measured 8 Mbps/1.8 Mbps at Penn Station. (Splash screen courtesy Klaus.)
A report from Canada's Globe & Mail indicates that Via Rail's Internet service could use more robustness, where cellular doesn't fully cut it. The firm that operates Via, 21net, only uses cellular connections on the Windsor-to-Quebec City route over which Internet is available. That's what the train operator wants. 21net recommends adding satellite for reliability.
Posted by Glenn Fleishman at 10:38 AM | Permanent Link | Categories: Rails | No Comments
Virgin Mobile is keeping its pricing tiers for its pay-when-you-need it broadband service, but bumping up included quantities of data: Virgin Mobile, now part of Sprint Nextel, has a unique 3G service, in which you can pay for limited amounts of data for limited periods of time. No contracts, and no other fees. And the division is upping the amount of data at each tier starting this morning.
Formerly, you could pay $10 to use up 100 MB within 10 days; with a 30-day usage period, you could pay $20 for 250 MB, $40 for 600 MB, and $60 for 1 GB. There are no overage fees because you are prepaying for a specific quantity of service. If you need more, you just buy another chunk.
The revised plan sticks with the $10/100 MB/10 days tier, but ups the data for each 30-day usage option: $20 gets you 300 MB (only a 50 MB increase), $40 gets you 1 GB (up 400 MB), but $60 now covers 5 GB or 400 percent more usage.
The $60 plan is identical in cost and data quantity to the contract-based 3G laptop service provided by AT&T, Sprint Nextel, T-Mobile and Verizon Wireless, but there's no commitment. (T-Mobile also allows you to pay full price for a USB modem and then pay on a month-by-month basis.)
I suspect Virgin bumped up these numbers because as a better deal it encourages more regular purchases without feeding out much more data. I suspect most people paying for 1 GB never reach that total, and that offering 5 GB won't encourage much more consumption relative to the jump in usage.
Just as carriers have all seemed to spawn prepaid, supercheap voice offerings--all fees are collected before usage--I expect we'll see more prepaid 3G, too. Postpaid plans are supposedly billed after the fact, but all the voice and 3G data contracts I know of required a month's advance prepayment of subscription fees, but allow you to run a tab for overages during the month of usage and then pay for those. Hardly postpaid, despite the definition of postpaid.
The plans all require the $100 Broadband2Go USB Device (a Novatel Ovation MC760 with microSD slot). The USB modem works over Sprint Nextel's network at EVDO Rev. A speeds where available, and supports several Windows flavors and Mac OS X 10.3 and later. (Some commenters on Virgin's open-answer FAQ say 10.4 is the minimum supported version.)
Posted by Glenn Fleishman at 4:31 AM | Permanent Link | Categories: 2.5G and 3G, Unique | No Comments
Netgear hasn't set the price of its new paired HD video adapters, but promises 99.9 percent reliability for multiple 1080p streams: The High-Performance Wireless-N HD Home Theater Kit, which will ship in Q3 2010, uses a 4-by-4 MIMO array to achieve the results Netgear claims, and I'm inclined to believe them. The company says it can push 40 Mpbs across multiple streams using these adapters, with the video sources being the Internet, IPTV systems, or other devices on the network. The adapters plug into Ethernet ports, and have a simple pairing mechanism.
The 4-by-4 array over 5 GHz coupled with the paired adapter method means that Netgear doesn't need to focus on throughput but coverage and consistency. The extra antennas let them use space-time block coding and other techniques to boost marginal signals and reduce errors in transmission. The device doesn't compress data, so the entire goal is to achieve sustained throughput.
Of course, Netgear could charge $500 for the pair, which would make them ridiculous, but I suspect a price closer to $250 based on how other devices have been marketed in the past, and the target audience for these products. A third adapter can apparently be used to extend coverage further.
Netgear also announced a May release at $79 of the 802.11n (2.4 GHz) Universal WiFi [sic] Internet Adapter, a driver-free Ethernet-to-Wi-Fi bridge that can be powered by a USB port even though no data is handled over USB (it saves a power adapter). The notion here is that instead of buying branded, proprietary adapters, you can just plug into Ethernet.
Posted by Glenn Fleishman at 3:34 PM | Permanent Link | Categories: Consumer Electronics, Streaming, Video | 1 Comment
Amtrak's promise last year of putting Internet service in Acela trains happened quite quickly: For a chronically underfunded government-like operation, Amtrak managed to get Wi-Fi installed fast in its Northeast high-speed Acela line after it said it planned to do so. The service, free for the interim, is in all 20 Acela trains. Amtrak has also made Wi-Fi service free in the six stations that serve Acela, and in Acela lounges. (The Wilmington Station will get unwired when renovations are completed in early 2011.)
Amtrak may wind up using Wi-Fi as yet another tool to bring passengers out of the air and onto the ground. With Wi-Fi at no cost in stations and on trains, the rail operator could use that as a marketing campaign against $7 to $10 per day airport and $5 to 13 per flight airline Internet access, where that's available.
Nomad Digital and GBS Group built the service out. Nomad has a many year history of train-Fi. Nomad and competitor Icomera are responsible for most of the train-based Internet access in the world. A few other firms have disappeared during that time or exited the business.
Posted by Glenn Fleishman at 1:19 PM | Permanent Link | Categories: Free, Rails | No Comments
A UK law under consideration and much reviled by privacy advocates would make independent Wi-Fi hotspots legally indefensible: The Digital Economy Bill is a particularly odious piece of legislation that attempts to enforce copyright by requiring ISPs to keep records and disconnect customers who engage in such acts.
This puts the government in the business of taking people off the Internet by enforcing actions for what would otherwise be civil violations, previously needing to be proved in court. Now, something approximating an assertion and a few letters could cause an ISP that doesn't respond appropriately to face huge fines and other troubles. A similar law in France was initially struck down as unconstitutional, but was modified lightly before being approved.
The reason for these laws is to keep media industries from engaging in publicity-adverse lawsuits against individuals, such as those the RIAA (Recording Industry Association of America) led against college students, children, and dead people before sputtering out and moving to this approach.
The not-quite-unintentional consequence of the UK law would, according to advice provided by an arm of the government, put undue burden on hotspots, libraries, and academic institutions. The law requires that most parties be either subscribers (end users) or ISPs; ISPs primarily provide access, and subscribers use it, although there are some fine points. In either case, copyright holders can notify ISPs of violations who are required to notify subscribers. After a small number of violations, the subscriber can be disconnected from any Internet service for some period of time.
If someone downloads an allegedly pirated video over a library, university, or hotspot could force that institution off line if it failed to meet specific notification terms; it's unclear how a hotspot could restrict a banned user without imposing high bars for access, whether free or fee. Larger operations could have login and credit-card verification requirements--used mostly as a way to block people instead of allow them.
Danny O'Brien of the Electronic Frontier Foundation previously raised this concern on 2 December 2009 in a post to the EFF's site: "The repeated demand by the entertainment industry that intermediaries should police their networks has been expanded by the bill to include the subscribers on the edge of the network. If you're not an ISP, but other people use your network to get their net access — if you run an open Wi-Fi spot, for instance, like the British Library — you'll now be vulnerable to being terminated or constrained by the actions of those users."
None of these efforts, of course, deters privacy. As Cory Doctorow, a UK resident and editor of BoingBoing, wrote about this issue:
"The Digital Economy Bill is being sold to us on the grounds that copyright infringement harms the British economy because of the importance of our entertainment industry. But while the measures in the DEB won't stop copyright infringement (copying isn't going to slow down -- as computers and the technology they enable gets cheaper and more widely distributed, copying will continue to speed up, just as it has done since the dawn of the computer industry), they will harm British business and British families, by making the Internet generally less useful and more difficult and more expensive for honest people to use."
In the US, we don't have such a law underway--as far as I'm aware--but media firms have struck deals with some ISPs (and some ISPs have refused) to engage in the same sort of behavior without government involvement.
Posted by Glenn Fleishman at 3:39 PM | Permanent Link | Categories: International, Legal | 1 Comment | No TrackBacks
Martin Beck has an enhanced attack against TKIP: One of the two researchers who brought us the TKIP Michael packet integrity attack has a refined technique. Beck's paper, "Enhanced TKIP Michael Attacks" [PDF download], describes how to work around certain assumptions in the MIC (Michael) checksum that's used to ensure a packet hasn't been tampered with to insert truly massive hunks of data without breaking a TKIP key.
For certain kinds of routine network traffic, enough data is already known in the right circumstances to brute force one missing piece and insert from 120 to 568 bytes, if I read the paper right. The Michael checksum isn't changed, but the packet is inserted as a fragment before a correctly checked hunk of data, so the receiver has no suspicion of tampering.
Worse, this technique can be used in some cases to decrypt data headed to the client, even though the TKIP key hasn't been recovered.
As with the previous attack, a lot of stars have to be in alignment. The biggest requirement is that TKIP has be the key type, not AES-CCMP. An attacker has to be proximate to sniff traffic and inject packets. The router has to be running Linux, like many Wi-Fi routers do. The router doesn't need to be compromised; there's a particular Wi-Fi packet sequence that's more predictable, and thus easier to use in the attack. Network QoS (802.11e/WMM) needs to be enabled as well.
If you can use the AES-CCMP key type (sometimes incorrectly called WPA2 by itself, but really the more advanced of two WPA2 methods), then you should! All corporations and other entities should already be using AES-CCMP, and with nearly all devices sold starting in 2003 supporting AES-CCMP, even home networks should be able to make that choice.
(A reader sent email asking if I wasn't mistaken: doesn't WPA2 only support the AES key? Yes, but it's also backwards compatible. If you have a WPA2 implementation, it may support TKIP unless the router maker has provided an option to lock into using AES only. Depending on the router, you might see "WPA, WPA/WPA2, WPA2" as a set of options, which corresponds to "TKIP, TKIP/AES, AES" for key types; or an explicit menu that lists key types after you select WPA or WPA2.)
Posted by Glenn Fleishman at 4:49 PM | Permanent Link | Categories: Security | No Comments
While T-Mobile's UMA offering has been around for years, Cablevision may be trying something new: Cablevision's COO mentioned in the company's earnings call today that the firm is testing phones that will switch seamlessly between cellular and Wi-Fi networks. That sounds an awful lot like UMA (unlicensed mobile access), a standard used for roaming by T-Mobile in the US and several carriers around the world.
T-Mobile offers UMA because it lets them leverage other companies' broadband and its network of home and roaming Wi-Fi networks. T-Mobile operates relatively few hotspots now compared to when it was anchored by Starbucks, but the key to UMA is voice over Wi-Fi over a wired or wireless broadband connection at home.
Cablevision has an even easier time of it, because it provides its CT/NY/NJ Wi-Fi network only to users that subscribe to its home cable broadband service. So any phone it offers can carry voice over Wi-Fi at home over its cable network and outdoors over its Wi-Fi network. In the past, Cablevision has partnered with Sprint for wireless service. However, I'm unaware of any production UMA gear that would work on Sprint's network.
The notion of UMA is to reduce the cost to a carrier of subscribers while providing subscribers with more unrestricted minutes. T-Mobile's plans, for instance, offer unlimited domestic calling over Wi-Fi even for plans with modest numbers of cellular minutes. This keeps customers loyal and happy with a lower cost structure. Femtocells have the potential to offer similar advantages to T-Mobile's competitors, but none are being priced or marketed in that way yet.
Posted by Glenn Fleishman at 2:57 PM | Permanent Link | Categories: 2.5G and 3G, Voice | 1 Comment