Verizon to pay $1 million over deceptive “unlimited” EVDO plans

The case of the limited "unlimited" EVDO has been settled: Verizon has agreed to pay out $1 million to customers that it has terminated for overuse of its high-speed data service. New York Attorney General Andrew Cuomo made the announcement today, saying that Verizon's decision came after a nine-month investigation into the company's services and marketing practices. The attorney general accused Verizon of producing misleading materials and deceptive marketing when it claimed that its data plans were unlimited. HangZhou Night Net

The issue came to light last year, when some customers found their accounts on the chopping block after downloading too much data over Verizon's wireless broadband service. The wireless provider prominently advertised its EVDO service as "unlimited," but the fine print indicated that it was only unlimited for certain things, such as e-mail and web access. Video, music, and other media did not fall into that category, and Verizon began enforcing an undisclosed bandwidth cap on users that the company decided downloaded too much.

Verizon eventually cut off some 13,000 customers for excessive use of its "unlimited" service, leaving those customers out in the cold with equipment that they could no longer use. But the Attorney General said that the service's limitations were not clearly and conspicuously disclosed, and that they "directly contradicted the promise of 'unlimited' service."

Verizon, which has voluntarily cooperated with the investigation since it began in April, has now agreed to reimburse the terminated customers for the costs associated with their now-useless equipment. Verizon estimates that it will come out to be about $1 million, with an additional $150,000 in fines paid out to the state of New York.

"This settlement sends a message to companies large and small answering the growing consumer demand for wireless services. When consumers are promised an 'unlimited' service, they do not expect the promise to be broken by hidden limitations," said Cuomo in a statement. "Consumers must be treated fairly and honestly. Delivering a product is simply not enough—the promises must be delivered as well."

The company has also agreed to revise its marketing of the wireless plans, in addition to allowing common uses of the broadband connection (such as video downloads). As you can see above, Verizon has already updated some of its marketing material.

Let your geek flag fly: a review of Eye of Judgment

A cold winter morning, in the time before the light

Eye of Judgment
Publisher: Sony Computer Entertainment
Developer: SCE Studios Japan
Platform: PlayStation 3
Price: $69.99
Rating: Teen HangZhou Night Net

A few days ago I had an experience of such intense geekiness that I felt like I was back in high school, in a garage somewhere, listening to a Led Zeppelin tape and playing Dungeon and Dragons. What caused this intense and oddly freeing embrace of the best parts of being a geek? I sat down with a good friend, set up Eye of Judgment, and played for hours upon hours. The cheap pot may have been traded in for a nice bourbon, but the overall feeling was the same.

Eye of Judgment is a game that doesn't just require you to be a geek, it demands that you give yourself over to the sheer insanity of the experience. You need a PlayStation 3, a nice television, the PlayStation Eye, a bunch of cards, a table near the PS3 so you can hook up the camera and lay down the mat, and, in an optimal situation, a friend sitting across from you.

When you're holding your deck of cards, staring into the eyes of your opponent before laying down that Biolith God that will lay waste to his entire force, you'll know that not only are you playing something that embraces many clichés of games you've played before while doing something completely new, you may have just reconfigured your entire room to do it.

What the players see on the left, what the television displays on the right

I lay my card down on the mat. The camera scans it, and on the television runes explode from the card's face before the deity emerges and destroys two of the four creatures my friend has laid down. He removes the "dead" cards from the mat, places them in his graveyard where the vanquished rot, and takes a sip of his drink. Only then do I realize I just blew all my mana, and he's been banking his for an equally impressive attack. He lays his card down, and our eyes turn to the screen to see what carnage his attacks will cause. Heavy metal music is playing is the background.

This is awesome. This is Eye of Judgment.

My Judgment. Let me show you it

Eye of Judgment is a little bit Magic: The Gathering and a little bit of the chess game from Star Wars. The new webcam from Sony looks down on the game mat, which features a 3×3 grid of squares. You lay down cards on the squares, summoning creatures to allow you possession of each slot. When you own five spots, you win. Sounds easy, doesn't it?

While you lay down cards on the mat, the actual play area may not look like much, but on your television you'll see a creature on top of each card, you'll see information on the hit points that each creature has left, as well as their attack strength and which direction they're facing. Each time you turn a creature, attack, or summon a new monster you use mana. You gain two mana points each round, and there are cards that allow you to gain more mana points by discarding creatures in your hand. Management of yourmana points is a major piece of strategy: do you summon a bunch of weak creatures early in the game or save your points to lay down the serious ordnance?

There is a spell card that costs no mana to cast and simply spins a character around. Since creatures can only attack in certain directions and spinning this particular creature cost six mana for every 90-degree turn, I effectively wiped out my opponent's mana pool in one swipe. The strategy is deep; you'll be learning new tricks and getting ideas for how to play your cards for a long time.

Give your living room to Eye of Judgment

Learning how to play and what the numbers and information on the cards mean takes a little bit of a time investment, although the printed manual and video-based tutorials make the process easier. Just a warning, the video tutorials aren't interactive, and they're incredibly boring. They actually explain to you that when someone brings cards over, they are the "owner" of those cards. I'm surprised they didn't explain that the cards were made of "paper" which is made from "pulped trees."

The game itself is not overly complicated. You can explain things and play a quick game in about an hour, and after that, players catch on rather quickly to the subtleties of strategy. Eye of Judgment could have been a complicated mess, but I'm happy to report the rules make sense and are easy to comprehend while not limiting play. Each battle is a series of questions. Doyou use mana now or later? Each square matches with an element, so ifyou play a water-based card on a water square you get a bump. Do you play a card now or wait for a better elemental slot to become available?

These are questions that will haunt you at night as you listen to Dragonforce and plan your next deck.

I've got a BA in mana management

How well the card game plays is a worthless thing to try to judge if the hardware doesn't work; if the camera has issues with the cards, then the whole set up is a waste. It can take a minute or two to calibrate the camera and to make sure the mat is lined up correctly, but the tools built into the game do a great job of making this painless. You'll also want to make sure the room you play in has plenty of light. After these two things are done, the camera does a fantastic job of seeing the cards and reacting quickly to your play.

This is made even more impressive by the fact that cards in play aren't the only things the camera has to keep track of. To attack or end your turn, you have to show the camera "action" cards to tell the game what you'd like to do, and this process is quick and easy. Sometimes the camera does lose track of the card, but when this happens you get a clear warning, and a slight adjustment is usually all that is necessary to make things right again.

We did bump into troubles once or twice with cards that the camera refused to recognize, but this was a very rare occurrence. Overall, the PlayStation Eye camera was more than up to the task of keeping play quick and fun.

My one complaint—and this will only become an issue in future games—is that even with good lighting the actual image the PlayStation Eye picks up is pretty muddy and indistinct. This won’t bother you in Eye of Judgment, where the screen is covered with overlays and you will only see your hand on the screen when you lay cards, but it doesn’t bode well for later games that will feature more video interaction. The picture is better than the last-generation Eye Toy, but the jump isn’t as large as I had been hoping for.

The camera stand breaks down quickly and easily for storage, and set up only takes a minute or so. You may want to flatten the mat when you first unpack it, as the little ridges from being packed are slightly annoying while laying down cards.

Comcast shooting itself in the foot with traffic shaping “explanations”

As the evidence that Comcast is doing something untoward with BitTorrent and other traffic on its network has mounted, the cable company has tried clumsily to fend off accusations of wrongdoing. The latest developments come in the wake of several conference calls held by the ISP in which it attempted to make a case for its practice of sending forged TCP reset packets to interfere with some P2P traffic. HangZhou Night Net

Timothy B. Lee, who is a regular contributor to the Tech Liberation Front blog as well Ars Technica, was invited to sit in on one of yesterday's conference calls, along with folks from a handful of think tanks. According to Tim, the Comcast engineer on the call said that the Lotus Notes problems were a known side effect of Comcast's traffic shaping practices, one the company was trying to fix. The engineer also "seemed to implicitly" concede that the accounts about the forged packet resets were accurate.

Delaying as a blocking tactic

The company still claims that it is isn't blocking BitTorrent and other P2P traffic, just "delaying it." In a statement given to Ars earlier today, a Comcast spokesperson denied that the company blocks traffic. "Comcast does not block access to any Web sites or online applications, including peer-to-peer activity like BitTorrent," the spokeperson told Ars. "Our customers use the Internet for downloading and uploading files, watching movies and videos, streaming music, sharing digital photos, accessing numerous peer-to-peer sites, VOIP applications like Vonage, and thousands of other applications online. We have a responsibility to provide all of our customers with a good Internet experience and we use the latest technologies to manage our network so that they can continue to enjoy these applications."

Comcast VP of operations and technical support Mitch Bowling put it this way. "We use the latest technologies to manage our network so that our customers continue to enjoy these applications. We do this because we feel it's our responsibility to provide all of our customers with a good Internet experience."

Another Comcast executive told the New York Times that the company "occasionally" delays P2P traffic, "postponing" it in some cases. His rather clumsy analogy was that of getting a busy signal when making a phone call and eventually getting through after several attempts. "It will get there eventually," is the takeaway message.

That's a distinction without any meaning. If someone is preventing my calls from going through and giving me a busy signal, the effect is the same. At the time I am trying to make the call, it's being actively blocked; calling it "delayed" is merely an exercise in semantics. Comcast is, in effect, impersonating the busy signal and preventing the phone at the other end from ringing by issuing TCP reset packets to both ends of a connection.

What's particularly troublesome is that Comcast's FAQ leaves customers with the impression that all content will flow unfettered through its network. One entry states that Comcast engages in "no discrimination based on the type of content," saying that the ISP offers "unfettered access to all the content, services, and applications" on the Internet. Another FAQ entry informs customers that Comcast does not "block access to any Web site or applications, including BitTorrent."

What did I do wrong?

Comcast's attempts to clarify its traffic shaping practices are having the opposite effect of what the company intends. As is the case with its nebulous bandwidth caps, customers can find themselves running afoul of what appears to be an arbitrary limitation imposed by the ISP. As a result, Comcast's customers don't really know that what they're paying for, aside from a fast connection that may or may not give them access to the web sites and applications they want. The company's public comments on the traffic shaping issue are intended to leave the impression that, like the bandwidth cap issue, this only affects a handful of bandwidth hogs. But judging by the comments we've seen from our readers and on other sites, there are either a lot more bandwidth hogs than Comcast leads us to believe, or the company's traffic shaping practices extend further than is being disclosed. Without some transparency from the ISP, we're left to guess.

Comcast has a handful of options to choose from. The company can own up to what it's doing and tell customers how to avoid running afoul of its BitTorrent regulations. Comcast could also continue on its current course, keeping its opaque traffic management practices in place. The cable giant's best option may be dropping the practice of sending false TCP reset packets altogether.

There are a couple of reasons that the third option may be the best choice for Comcast. First, it may be against the law. An Indiana University PhD student and Cnet contributor believes that the illicit reset packets may violate state laws in Alabama, Connecticut, and New York against impersonating another person "with intent to obtain a benefit or to injure or defraud another" (language from the New York law). In sending out the spoofed packets, Comcast is impersonating the parties at either end of the connection.

When the market can't sort things out

Legal concerns aside, Comcast is providing net neutrality advocates with plenty of ammunition. Comcast is not running a neutral network right now, and its traffic shaping choices are degrading the broadband service of many a Comcast customer.

In a perfect free market, customers would be free to pack up in leave Comcast for greener and more open broadband pastures, but the competitive landscape in the US doesn't always provide that kind of choice. More than a few Comcast customers are faced with the choice of Comcast or dial-up, leaving them with the Hobson's choice of hoping their data packets can evade Comcast's traffic shaping police or not having broadband service at all.

KDE development platform appears

HangZhou Night Net

This Wednesday in KDE land sees the start of the Release Freeze for KDE 4.0, anticipating a final release in December. By then, KDE 4 will have been in development for two years and five months, counting from the aKademy conference in Malaga, Spain, in July 2005.

This freeze is significant for KDE development as evidence of an increasing professionalization in the KDE release process. Where past major releases like KDE 2 (2000) and KDE 3 (2002) involved a relatively close-knit group of insiders releasing a set of source tarballs for Linux distributions to package, the KDE landscape in 2007 is broader and much more diverse. Seventeen modules made up KDE 3.0, and that contained pretty much the entire desktop and all the software that most users would run on it.

During the lifetime of KDE 3, development has become decentralized and some of the most popular applications, such as Amarok, Kopete and KDevelop emerged. They formed their own developer communities who may have little overlap with the core group of KDE library and platform hackers.

In addition, groups of businesses such as the Kolab Konsortium (we kid you not, those Ks are authentic German) have sprung up to serve the needs of companies using KDE. Their ‘Enterprise’ Branch’ flavor of the Kontact PIM suite, featuring additional enterprise features and QA work, is released on its own schedule, independently of the core KDE modules. Completing the scene, a swarm of smaller projects and individual developers take advantage of the KDE libraries, producing thousands of applications, as seen at www.kde-apps.org.

This expansion has caused major changes in the KDE release process. The era of the lone release manager (represented in the past by David Faure and Stephan Kulow) bearing responsibility for getting every module ready just in time for distributions’ release dates has ended, and instead a Release Team representing KDE’s various stakeholders oversees the readiness of all its components.

Also, the release itself has become staggered. To give all the projects and developers who rely on the KDE libraries time to adapt to the changes of a major release, these core libraries and runtime components are being frozen ahead of the rest of the desktop and its applications.

On October 30, the KDE Development Platform version 4.0.0 will be released. This forms a stable set of packages including the libraries, the PIM libraries, and runtime components of the base desktop such as the help browser, IO slaves, and the notification daemon that third party developers require to port their apps to KDE 4. They will be able to download this platform as binary packages for several distributions, easing the porting process.

KDE is expanding the scope of its ambition with this release, by targeting developers of scripting languages as well as its traditional C++ constituency. By shipping bindings for Python and Ruby as part of the Development Platform, the KDE project hopes to signal that these languages are fully accepted for KDE application development. Bindings for C# are due to follow shortly afterwards. And for the first time, KDE will be released for Mac OS X and Windows.

Targeting two new operating systems poses a challenge to the community’s resources. A brief poll of developers showed that although ‘most things work’, the KDE Development Platform version won’t be ready for these platforms at the same time as the Linux flavor. It is hoped that in the long term, this investment will pay off by bringing more participants to Free Software.

The significance of this milestone to KDE fans and users is that the rest of the applications making up the core KDE Desktop are now soft-frozen. While third party developers are just getting to grips with KDE 4, the applications that form the central KDE experience need to get in shape for the KDE 4.0 release. With a lot of issues facing KDE hackers before 4.0 is a usable desktop, all work on new features and UI is stopped, and efforts focus on fixing the inevitable, long list of bugs. A state of the user desktop is the topic of an upcoming article.

Hands on with Google’s OCRopus open-source scanning software

The first official alpha version of Google's OCRopus scanning software for Linux was released yesterday. OCRopus is built on top of HP's venerable open-source Tesseract optical character recognition (OCR) engine and is distributed under the Apache License 2.0. HangZhou Night Net

OCRopus uses Tesseract for character recognition but has its own layout analysis system that is optimized for accuracy. The OpenFST library is used for language modeling, but it still has some performance issues. OCRopus is designed to be modular, so that other character recognition and language modeling components can be used to eventually add support for non-Latin languages. An embedded Lua interpreter is used for scripting and configuration. The developers chose Lua rather than Python because Lua is slimmer and easier to embed. This release also includes some new image cleanup and de-skewing code.

We tested OCRopus with several kinds of content, including scanned documents and screenshots of text. The accuracy of the output varies considerably depending on the quality of the input data, but OCRopus was able to provide readable output in about half of our tests. Several of our test documents caused segmentation faults; this seems to occur when there is extra page noise or the letters aren't consistent.

We observed several common errors. For instance, the letter "e" is often interpreted as a "c" and the letter "o" is often interpreted as a "0" in scanned documents. OCRopus provides better results when scanning text that is printed with a sans serif font, andthe size of the font also has a significant effect on accuracy.

The following examplesshow the typical output quality of OCRopus:

Tpo' much is takgn, much abjdegi qngi tlpugh we arg not pow Wat strength whipl} in old days Moved earth and heaven; that which we are, We are; QpeAequal_tgmper of hqoic hgarts, E/[ade Qeak by Eirpe ang fqte, lgut strong will To strive, to Seek, to hnd, and not to y{eld.

Tho' much is taken, much abides; and though We are not now that strength which in old days Moved earth and heaven; that which we are, we are; One equal temper of heroic hearts, Made weak by time and fate, but strong in will To strive, to seek, to find, and not to yield

The installation process is relatively straightforward for experienced Linux users, but there were a few bumps. I tested OCRopus on my Ubuntu 7.10 desktop computer. The only dependency that isn't in the Ubuntu repositories is Tesseract, which I had to download separately and compile. Since I already hadmany development packages installed, the only dependencies I needed (apart from Tesseract) were libtiff-dev and libaspell-dev. If you don't have the aspell dependency installed, OCRopus will still build, but it will trigger an error saying that the word list is unreadable.

I also ran into a snag with Tesseract, which apparently only comes with placeholders for the language data files, causing OCRopus to emit an error saying that the unicharset file is unreadable. In order to resolve that problem, I had to download the English language data files and decompress them into /usr/local/share/tessdata.

According to the release announcement, a beta release is planned for the first quarter of 2008. The focus for the beta will be on better performance and output quality, whereas the focus of this release was functionality.

Google's involvement in the project is motivated by the company's interest in digitizing printed documents. Open-source OCR technology could be valuable in many other contexts as well. Government agencies that want to digitize paper records, for instance, could one day benefit from OCRopus. Although OCRopus is weak in many areas, it has some real potential.