Wearable technologies from 35 exhibitors hailing from 10 countries were on display at the fourth annual SIGGRAPH Cyber Fashion Show this week. Sensor-equipped purses from Gauri Nanda of MIT were designed to communicate with skirts and scarves so that the wearer can be alerted to inclement weather, the location of misplaced items, and other daily reminders. Emagin and Total Fire Group developed a head-mounted thermal camera system to help firefighters locate flame sources under conditions of poor visibility, while Japan's Wearable Environmental Information Networks (WIN) spotlighted Report-the-World, a trench coat augmented with concealed cameras for recording 360-degree panoramic images, a small computer, a ring-embedded speaker for transmitting location-based audio instructions, and a head-mounted display. Another WIN product at the fashion show was Dog @ Watch, a plushy children's device worn on the wrist that features a GPS sensor, a cell phone for contacting parents, and an alarm sensor to monitor the wearer's safety. Some of the featured items were more whimsical than practical, such as attire by Keio University's Akira Wakita that displayed luminous codes indicating body temperature. Electroboutique showed off virtual reality goggles that "paint" one's surroundings with Photoshop-like filters.
Click Here to View Full Article

A cry for dramatic reform to the U.S. patent system is echoed by lawyers, lawmakers, inventors, and companies, but few can agree on what kind of remedy would address the system's many problems, which include an overwhelming backlog and dubious criteria for patentability. Rep. Lamar Smith (R-Texas) warned in a statement that innovation would be stifled without "adequate protection" for intellectual property, and this would have a negative effect on job growth and the national economy. Legal disputes about patents are running rampant, a trend partly driven by the ambiguity of the current definition of patentable inventions as those that are new and non-obvious to a person "having ordinary skill in the art." Brandeis University dean of arts and sciences Adam Jaffe believes patents should only be awarded to "truly new and non-obvious" inventions. Suggested reforms include Rep. Smith's Patent Reform Act of 2005, which would make the acquisition of patents more difficult and reduce the amount of damages owed to a plaintiff in a patent infringement lawsuit, as well as grant the courts more discretion in authorizing injunctions. Other proposed reforms are designed to reduce the likelihood of litigation by concentrating on the patent handling process: Some reformers would like amendments requiring companies to generate products from their patents or at the very least license the patents for others to implement practically. Non-U.S. patent systems are also serving as an inspiration for reformers--the European Parliament, for instance, has roundly rejected the concept of software patents. Another key differentiator between the U.S. patent system and others is the definition of an inventor: In the United States, the inventor is the person who creates a product, while the rest of the world defines an inventor as the first person to file an application with the patent office.
Click Here to View Full Article

The mainframe is enjoying a comeback of sorts with a resurgence in centralized IT functions among many organizations; also contributing is the need for companies to replace retiring mainframe experts and Chinese, Eastern European, and other international companies investing in mainframes to acquire additional computing power. Murray McBain with the Royal Bank of Canada says mainframes support a significant percentage of business for 95 percent of Fortune 1,000 companies, while the Clipper Group's Mike Kahn reports that mainframe work offers job security, a key consideration for many computer science students. IBM is moving ahead with its plan to churn out 20,000 mainframe-trained employees by the end of the decade through collaborations with corporations and educational institutions under the auspices of its Academic Initiative. Kahn says many computer science graduates lack large-scale project experience and the skills to understand enterprise computing, as both they and their teachers have been trained on PCs. He observes that mainframe programs such as IBM's are particularly successful at community colleges and night schools, which are more firmly committed to graduating employable students than some of the more first-class schools. Academic Initiative program manager Mike Bliss notes that collaborating with smaller schools and community colleges can be easier because such institutions are less specialized, less bureaucratic, and can implement classes faster. There is a serious decline in student computer science enrollments in North American schools, and Kahn says the dot-com meltdown and the media's focus on outsourcing have contributed to this trend. McBain recommends playing up computer science's challenging or exciting aspects in order to attract more students.
Click Here to View Full Article

VeriSign's Aristotle Balogh believes the hacker wars reached a critical milestone last year as hackers transformed from irritating nuisances into serious threats. An Aug. 2 report from IBM estimates that over 237 million security attacks occurred worldwide in the first six months of 2005, a 50 percent increase over the same period last year. Moreover, "customized" attacks targeting specific individuals and companies rose dramatically, which Balogh said is indicative of hackers' expanding skills. Not only have attackers' methods become more sophisticated, but their motivation has shifted toward profit-making and their activities are now supported by the criminal element, according to IBM analyst Jeremy Kelley. VeriSign has been channeling vast sums of money and computational power into security improvements: The company now keeps multiple copies of the "A" DNS root server on various computers throughout the globe, instead of having just one copy on a single system. In addition, the main "A" root computer's location is unknown to all but a few employees. Another enhancement is additional computing capacity to prevent the disablement of VeriSign's network by distributed denial of service attacks, an important safeguard since the network currently manages between 12 billion and 13 billion look-ups for Internet addresses per day. Balogh said hackers will have a harder time altering address records stored by ISPs thanks to revisions to the Internet domain system currently underway, but he pointed to the emergence of new hacking strategies, including "zero-day" attacks that exploit security flaws on the day of publication, and criminals' practice of leasing networks of hacked home PCs to one another.
Click Here to View Full Article

University of Washington computer-science program director David Notkin, who recently joined the board of the Computer Research Association, suggests that IT jobs are more plentiful now than they were prior to the dot-com boom--at least for people with design and other higher-level computing skills. The surge in competition means that possessing just average computing skills is no longer enough to guarantee a job, according to New York Times columnist Thomas Friedman. Notkin says the false assumption that IT jobs are scarce since the Internet bubble's implosion is partly responsible for a decline in the number of students studying computer science. He is attempting to encourage more people to pursue computer science by courting female and minority students, an effort complicated by persistent stereotypes of IT workers as socially maladjusted Caucasian males working in an isolating environment. Notkin says person-to-person interaction is an important element of programming, which means good social skills are a must. Computing is also a challenging and meaningful area of study that is applied to nearly every field. UW students, for example, are developing programs to assist people who suffer from Alzheimer's.
Click Here to View Full Article

The House and the Senate will hammer out a compromise on the Pentagon's fiscal 2006 authorization bill in September; both measures would provide more than President Bush sought for its science and technology programs. The House bill sets aside $892 million more for science and technology, while the Senate measure allots $482 million more for such programs. The House bill provides a total of $11.4 billion for science and technology, including $160 million for advanced technologies. The House Armed Services Committee took issue with the reduction of funding for science and technology programs in the long-term budget forecast. "The committee cannot emphasize too strongly the need for the department to maintain a strong and robustly funded science and technology program that will provide the advanced technologies needed to assure technical dominance of U.S. armed forces on any current or future battlefield," the committee report on the bill said. The Senate seeks to permanently mandate the science, mathematics, and research for transformation (SMART) program in its measure.
Click Here to View Full Article

Concealed Internet sensors designed to spot malware before it becomes a threat could be thwarted by future computer worms if preventive action is not taken, according to two papers presented at the Usenix Security Symposium on Aug. 4. One paper authored by University of Wisconsin computer scientists detailed how attackers could exploit sensor maps to determine the location of hidden Internet sensors, which could then be circumvented so that criminal activities could proceed unobserved. Co-author and presenter John Bethencourt said an attacker could simply send data packets to IP addresses and check to see whether such activity was documented on the sensor reports, a technique that would work "even with reasonable constraint on bandwidth and resources." The Wisconsin researchers said encryption, IP address concealment, and other current countermeasures are inadequate, and proposed that pervasive IPv6 adoption could help shore up defenses by extending IP addresses. Japanese researchers reported in the other paper that sensor attackers do not need a full list of sensor addresses to find the sensors, and they developed several algorithms that can locate the sensors in a unnervingly brief time. Both studies agreed that reducing the level of detail in the sensor networks' public reports would help mitigate the threat.
Click Here to View Full Article

A new technique with the potential to increase the security of password authentication has been developed by Stanford computer science researchers, who created a free and open browser plug-in that helps shield passwords. Stanford student Nicholas Miyake, one of the researchers who developed the extension, says many people use the same password to access multiple Web sites, which means that hackers could steal passwords from sites with low security and use them to penetrate more sensitive sites. Instead of storing passwords securely, the PwdHash program generates new passwords designed to fool hackers: The researchers report that when PwdHash users visit a site that needs a password, they either select a password that begins with the "@@" prefix or push a special password key; PwdHash then scrambles the actual password and combines it with text from the site's address, and this jumbled code is what hackers will see instead of the real password. PwdHash's security method can also protect against phishing, as phishers who set up bogus Web sites to lure the unwary will unknowingly steal false passwords that are useless at other sites. PwdHash plug-ins for Internet Explorer and Mozilla Firefox are available, as is a Web-based version. PwdHash does not thwart every hacker strategy for stealing passwords, and users say certain Web sites and browsers are incompatible with PwdHash. The software's source code was deliberately left open and free in the hopes that "someone bigger than us with more marketing muscle will pick up the project and put it in the hands of the average user," says Stanford computer science doctoral student Colin Jackson.
Click Here to View Full Article

The San Diego Supercomputer Center's Karan Bhatia, SICS' Per Brand, Lawrence Berkeley National Laboratory's Karlo Berket, Microsoft's Alex Mallet, and Oracle's Sergio Mendiola have authored a new Global Grid Forum paper discussing how grid computing and peer-to-peer (P2P) applications can be made interoperable. The document, "Peer-to-Peer Requirements on the Open Grid Services Architecture Framework," states that it is the P2P community's responsibility to ascertain how next-generation Web services and OGSA-oriented protocols can be used to support P2P applications. The authors admit that P2P systems and traditional server-based grid systems have dissimilar security and trust models, usage models, and connectivity properties. However, they argue that the vast presence of available desktop systems demonstrates the value of merging desktops and servers into a single grid system. The authors believe ubiquitous network links and interactions between devices, services, systems, organizations, and people will become the norm, and a complex global infrastructure will be needed to accommodate the many applications supporting these interactions.
Click Here to View Full Article

A directive supported by the European Commission calls for the criminalization of "attempting, aiding, or abetting and inciting" copyright infringement--a proposal that could, like the U.S. Supreme Court's recent ruling in the case of MGM v. Grokster, make developers of file-sharing software indirectly liable for digital piracy in certain EU member countries. University of St. Gallen law professor Urs Gasser warns that innovation could suffer, as the directive would make software and IT firms think twice about developing technologies that pirates might later adopt. He also says legislative interpretations of the directive could vary across EU member states, which would exacerbate uncertainty over what inventions can be deemed lawful or unlawful; more precise classifications of legal and illegal, particularly as they pertain to software and IT, are necessary, according to Gasser. The adoption of the directive could be helpful to U.S. firms that wish to pursue litigation against indirect copyright infringers in Europe. Among the proposal's suggestions is the establishment of a four-year maximum prison sentence and a 300,000-euro fine for a "criminal organization" that commits piracy, as well as the harmonization between differing European anti-piracy criminal penalties by requiring EU member states to incorporate consistent maximum prison sentences into their sentencing guidelines. Thomas Dillon, legal counsel for the Motion Picture Association, says making indirect acts of copyright infringement illegal is merely gloss. "The armory of the directive is that they have gathered together the different ideas of the member states of how you impose remedies, which are not uniform across the union," he says.
Click Here to View Full Article

For information on ACM's activities regarding MGM V. Grokster, visit http://www.acm.org/usacm.

Hackers are utilizing Google's massive database, with blame falling to Internet users who are unaware what Google makes available, according to Computer Sciences Corporation researcher Johnny Long. Recently, Long and other researchers were able to harness Google to find an unprotected Web interface to someone's household electrical network, an exposed printer network, private branch exchange enterprise phone systems, routers, Web cameras, and other devices. Many hackers are using Google to search company networks without drawing the attention of company security staff. The information hackers are acquiring through this method usually takes the form of "Google Turds," which is Long's word for seemingly nonsensical data. Also, knowledgeable Google users can enter certain searches to discover SQL passwords, SQL error data, and other sensitive information. This is often done through a combination of Google queries and text processing tools. Google is not to blame for the issue, but it has recently started preventing some attacks by rejecting certain queries.
Click Here to View Full Article

A team of graduate students at Carnegie Mellon University has developed a radio program called Roadcasting that enables users to share music with listeners from other cars up to 30 miles away; the software also learns a user's tastes and offers the capability to search for music of a particular genre across all available stations. Filtering, the tool Roadcasting software uses to match content with a user's preferences, has become increasingly popular at sites such as Amazon and Netflix, where recommendations are generated based on past selections. Jim Garretson, one of the system's developers, believes his program will help to bring people together and form a community of listeners with common interests. The students were commissioned by an unidentified auto maker to develop the program with a target availability date of 2010. Roadcasting will also tap into podcasts, the popular amateur radio programs that the software would recommend based on preference and history. Since the software invites others to share in a user's musical programming, it has incorporated security measures to conceal a user's identity. Before the technology hits the market, developers will have to refine certain aspects, such as licensing issues surrounding the distribution of content, though the program will not offer downloading capability.
Click Here to View Full Article

Schools such as MIT and Stanford University are embarking on projects to build digital libraries where access to knowledge is just a mouse click away. Many of these initiatives were jump-started by the U.S. government's plans to digitize 95 percent of its material this year, while Google's project to create vast repositories for digitized versions of some of the world's largest knowledge collections raised the stakes even higher. Stanford is planning a bookless engineering library comprised of group study rooms, a communal workspace, and computer terminals providing access to the Web as well as millions of documents; librarians will train students to use heuristic techniques for finding information. The school is also developing search technologies that use statistics or taxonomy to assess results. "We're really in a period of challenging transition, where we sort of know how to provide digital access to information, but we're very concerned about how to build a scholarly record over the long term," notes MIT library official MacKenzie Smith. MIT's digital library projects include the D Space repository; the use of open-source Lockss software from Stanford to cull electronic journals from the Web for archival; and a collaboration with print and online publishers to acquire long-term access to digital materials that may be under subscription and that will ultimately be removed from the Web. Many book digitization efforts involve rigorous research into the rights of works and securing permission to use them because of changes to copyright laws, which can differ internationally. "The library that acts as a steward will have to learn what it means to capture and persistently manage new vehicles of information," says Daniel Greenstein with the University of California's Digital Library project.
Click Here to View Full Article

LABELS is an IST-backed project that aims to optimize the capacity of existing fiber-optic cables in an effort to help address the growing demand for faster data transmission. LABELS is exploring ways to surmount logjams in the movement of data; the optical method the researchers are advancing allows data to be transmitted through multiple wavelengths, whereas electrical systems are limited to one bandwidth. Subcarrier multiplexing and the method of swapping labels in transferring packet data allow nodes to modulate wavelengths with greater flexibility than traditional wavelength division multiplexing (WDM) techniques, which tend to lock signals at set wavelengths. The researchers claim their method is ideal for packet transmission, where they see shortcomings in existing WDM. Early tests of the LABELS program have yielded speeds of 20 Gbps, and the researchers have their sites set on attaining 40 Gbps; LABELS is also developing a technique to enhance radio frequency processing by substituting optical filters for the current electronic ones to net more bandwidth. The optical filters operate at frequencies between 130 MHz and 20 GHz, and have the potential to reach 60 GHz. Telefonica I+D, a partner in the project, will test the RF technique this year for switching channels in UTMS, which has a low frequency. Valencia Technical University researcher Jose Capmany estimates that the technology could be market ready by 2010.
Click Here to View Full Article

The Khronos Group consortium announced OpenGL ES 2.0, a new iteration of the OpenGL desktop application programming interface (API), at the SIGGRAPH 2005 conference on Aug. 1. OpenGL ES, designed for use in cell phones, handhelds, and embedded systems, offers more flexibility than its precursors through the reduction or elimination of some of the older "fixed function" libraries, while Khronos Group Chairman Neil Trevett of nVidia said the most important tweak was the removal of redundancy. He projected that as many as 800 million OpenGL ED graphics-enabled cell phones could be shipped by year's end, increasing to an annual rate of 1 billion units. Small hardware's growing capability, combined with surging demand from wireless service providers for snazzier applications, illustrates a desire among independent software developers for a single graphics platform that supports a framework for applications that can be ported to various brands and deployments. Trevett said a program can draw a polygon using four distinct techniques because of the need for the broader OpenGL API to exhibit downward compatibility; OpenGL ES uses the latest polygon-drawing method to eliminate redundancy and reduce implementation costs while retaining functionality. Functions that were originally created for high-performance devices can scale down to lower-performance devices without changing their names or input constraints, thus making small hardware speedier and more power-efficient without major performance tradeoffs. The high-performance value of OpenGL ED will be tested through its use with the Sony PlayStation 3 game console.
Click Here to View Full Article

In an interview at the annual Microsoft Research Faculty Summit, Microsoft Chairman Bill Gates said he expects higher education to change dramatically as a result of new technologies. He predicted the continued enhancement of tablet PCs and their eventual ascendance to the preeminent computer on college campuses; a Microsoft software architect demonstrated a next-generation tablet with a display that partially separates from the unit and features a nesting keyboard. Gates restated his concern about students' waning interest in computer science as well as declines in federal funding in science and engineering research, which could be detrimental to America's economic growth. Responding to questions from attendees on how the fall-off in computer science students can be reversed, the Microsoft CEO said his company's best approach "is just share examples of what kind of jobs" are available, and their interesting aspects. Among current and upcoming Microsoft offerings discussed at the conference was Microsoft Office 12, which one company official said would allow peer-to-peer-based file sharing and document revision; and Conference XP, which facilitates remote, real-time conferencing over the Internet. Gates assured attendees that Microsoft is effectively addressing software security issues, insisting that security is the company's primary R&D area and that the forthcoming Windows Vista iteration will have major security improvements. Windows Vista was touted in another conference session for its ability to more efficiently permit users to configure, browse, and search for data. When confronted with the possibility of colleges migrating to free, open-source products, Gates cited Office's popularity in academia as a result of its continued augmentation and affordable price.
Click Here to View Full Article
(Access to this article is available to paid subscribers only.)

Many experts say that new 3D technologies could be the catalyst for the rollout of a digital cinema standard that promises to reduce film distribution costs and re-invigorate the experience of going to the movies, which translates into healthy profits for studios and theaters. Disney's announcement that it will outfit 100 theaters in 25 markets with digital 3D equipment in time for the release of a new animated feature in November is likely to prompt other studios to move similar initiatives forward, according to RealID CEO Joshua Greer. "Disney drew a line in the sand, and now everyone will have to jump in," he says. Defining a standard for mastering, distributing, and projecting digital 3D is the job of a new work group at the Society of Motion Picture and Television Engineers (SMPTE). There are several approaches to digital 3D cinema: RealID aligns two synchronized images via a circular polarization technique that is applied to the image with a special lens attachment on the projector, while viewers wear disposable glasses; they can angle their heads as much as 35 degrees before the image starts to ghost, but most theaters do not carry the silver-coated screen the RealID approach requires. In-Three's method eliminates ghosting and the need for silver-coated screens through the use of active-shutter glasses controlled by an infrared beam synched up to the projector. The SMPTE's 3D specification could call for alteration of the JPEG2000 codec underlying the studios' digital 3D standard, and suggest refresh rates that exceed the minimum 48 Hz of two 24 frame per second streams to remove flicker. The expanded storage and interface requirements that could result from the second implementation could make a new server-to-projector interface necessary.
Click Here to View Full Article
(Access to this site is free; however, first-time visitors must register.)

The U.N.'s Working Group on Internet Governance (WGIG) wants control of the Internet to pass from the U.S. Department of Commerce to an international coalition of governments, companies, and civilian entities. The Commerce Department holds the keys to the Internet through the Domain Name System's (DNS) master root server, which is based in the United States; this effectively gives America absolute authority over changes to the master root server file. WGIG executive coordinator Markus Kummer says this is worrisome for some WGIG members, since the United States could theoretically abuse this authority. The nonprofit ICANN manages the DNS system on the Commerce Department's behalf, and Kummer credits ICANN with its solid administration of international agreement on DNS amendments. However, the ad-hoc consensus process is problematic, and the global reach of the Internet has provoked the governments of Third World countries and elsewhere to demand a forum for discussion and action on numerous Internet policies. The WGIG believes U.S. control impedes efforts to improve the Internet, including initiatives to make Net access more affordable to developing countries, and the creation and enforcement of anti-cybercrime and privacy-protective measures. Proponents say only a multinational forum can instill fairness in non-DNS issues. The WGIG proposed recommendations for changing the Internet governance structure at a meeting last week, and these proposals will be voted on at the upcoming World Summit on the Information Society; among the suggested changes is the creation of an all-encompassing Internet policy authority that would run ICANN, and enhancing the roles of ICANN or its government advisory committee (GAC).
Click Here to View Full Article
(Access to the full article is available to paid subscribers only.)

User interface designer and author Niall Murphy insists that user errors must be considered in the design of a graphical user interface (GUI) if the end product is to provoke less frustration. He writes that reducing the number of error messages an interface produces requires the elimination of those errors in the design process; otherwise, the user will view the interface as less reliable, even if the failures are not caused by software bugs or technical malfunctions. Error messages that chastise the user can be a major turn-off, and Murphy recommends that such messages be reworded or made context-sensitive so that the user can follow a clear path to the solution without feeling embarrassed, humiliated, or insulted. An interface designer must avoid the employment of loud or intrusive noises to signal errors, since they can also be a source of embarrassment, especially for office workers. Murphy maintains that error messages should be designed to specify the type of error, the reason for its occurrence, and what remedial action should be taken. A designer is also responsible for making a distinction between errors caused by unintended actions or slips, and those caused by bad decisions. Murphy notes that the effectiveness of confirmation messages and warnings can be undermined by capture errors: For instance, the value of a confirmation request can be lost if the message appears frequently, prompting the user to automate his response. Another common error is one of recognition, such as when a GUI presents data with two different purposes in a highly similar manner, leading the user to make the wrong decision.
Click Here to View Full Article