Linux fréttir

An anonymous reader quotes a report from ZDNet: Google has raised a handful of concerns with Australia's pending Telecommunications Legislation Amendment (International Production Orders) Bill 2020 (IPO Bill), including the Commonwealth's choice of phrasing, the avenues proposed for record-sharing, and the Bill being at odds with the purpose of the United States' Clarifying Lawful Overseas Use of Data Act (CLOUD Act). [...] In a submission [PDF] to the Parliamentary Joint Committee on Intelligence and Security (PJCIS) and its review of the IPO Bill, Google said while it encourages and supports efforts by the Australian government to negotiate an executive agreement, it said there are certain elements of the Bill that give it cause for concern. "Especially when considering how the interception powers under this Bill could be used in tandem with technical capability notices under the controversial Telecommunications and Other Legislation (Assistance and Access) Act," it wrote. Making a recommendation to the PJCIS, Google said the Bill should not apply to service providers in their capacity as infrastructure providers to corporations or government entities, saying corporations or government entities are best placed to produce the requested records themselves. Under the Bill, designated communications providers are instructed to provide any requested communications and data to the requesting agency or the Australian Designated Authority. Google would prefer the authority to be a two-way channel. Google also poked holes in the Bill's enforcement threshold. Civil penalties for non-compliance with an IPO establishes a framework for compliance. If a designated communications provider receives a valid IPO and the designated communications provider meets the "enforcement threshold" when the IPO is issued, the designated communications provider must comply with the IPO. Google labelled the two-step test that is the threshold, a "relatively low bar to meet." "Failure to comply with an IPO may lead to a civil penalty of up to AU$10 million for body corporates. The imposition of a mandatory obligation to comply with an IPO is contrary to the purpose of the CLOUD Act which is to lift blocking statutes, but explicitly does not create a compulsory obligation on service providers," it said. Specifically, the search giant said it was concerned by the attempt to impose a mandatory obligation on overseas-based designated communications providers that exists "only in the construct of an otherwise non-compulsory international agreement." Google is seeking further information about the role that eligible judges will play in approving IPOs that involve the interception of communications. It also wants the appeal options contained within the Bill to be strengthened.

Read more of this story at Slashdot.

New submitter penandpaper shares an excerpt from a NASA press release: NASA has selected three U.S. companies to design and develop human landing systems (HLS) for the agency's Artemis program, one of which will land the first woman and next man on the surface of the Moon by 2024. NASA is on track for sustainable human exploration of the Moon for the first time in history. The human landing system awards under the Next Space Technologies for Exploration Partnerships (NextSTEP-2) Appendix H Broad Agency Announcement (BAA) are firm-fixed price, milestone-based contracts. The total combined value for all awarded contracts is $967 million for the 10-month base period. The following companies were selected to design and build human landing systems: - Blue Origin of Kent, Washington, is developing the Integrated Lander Vehicle (ILV) -- a three-stage lander to be launched on its own New Glenn Rocket System and ULA Vulcan launch system. - Dynetics (a Leidos company) of Huntsville, Alabama, is developing the Dynetics Human Landing System (DHLS) -- a single structure providing the ascent and descent capabilities that will launch on the ULA Vulcan launch system. - SpaceX of Hawthorne, California, is developing the Starship -- a fully integrated lander that will use the SpaceX Super Heavy rocket. "With these contract awards, America is moving forward with the final step needed to land astronauts on the Moon by 2024, including the incredible moment when we will see the first woman set foot on the lunar surface," said NASA Administrator Jim Bridenstine. "This is the first time since the Apollo era that NASA has direct funding for a human landing system, and now we have companies on contract to do the work for the Artemis program." Further reading: SpaceX and NASA Break Down What Their Historic First Astronaut Mission Will Look Like

Read more of this story at Slashdot.

Whitehat and blackhat next?

The British government's computer security gurus have announced they will stop using the terms whitelisting and blacklisting in their online documentation.…

Pioneering Australian scientist Robert May, whose work in biology led to the development of chaos theory, has died at age 84. The Guardian reports: Known as one of Australia's most accomplished scientists, he served as the chief scientific adviser to the United Kingdom, was president of the Royal Society, and was made a lord in 2001. Born in Sydney on January 8, 1938, May's work was influential in biology, zoology, epidemiology, physics and public policy. More recently, he applied scientific principles to economics and modeled the cause of the 2008 global financial crisis. On Wednesday, his friends and colleagues paid tribute to a man who they said was a gifted polymath and a "true giant" among scientists. Dr Benjamin Pope, an Australian astrophysicist and student at Oxford from 2013 to 2017, said May was a role model, and meeting him was a highlight of his university career. "I became aware of his achievements almost as soon as I learnt anything about physics in university," Pope told Guardian Australia. "My first contact with computer programming was at the University of Sydney, in first year physics, where the example is to recreate Robert May's experiment with the bifurcation diagram and the logistic map. "His bifurcation diagram is one of the iconic diagrams in physics," he said. "[And] he made what was between three or four independent discoveries that lead to chaos theory. You might have heard of the butterfly effect ... May's is probably the other foundational, computational model of chaos."

Read more of this story at Slashdot.

An anonymous reader quotes a report from IEEE Spectrum: The city of Carmel, Ind., has trucks for plowing snow, salting streets, and carrying landscaping equipment. But one cherry-red pickup can do something no other vehicle can: produce its own hydrogen. A 45-kilogram metal box sits in the bed of the work truck. When a driver starts the engine, the device automatically begins concocting the colorless, odorless gas, which feeds into the engine's intake manifold. This prevents the truck from guzzling gasoline until the hydrogen supply runs out. The pickup has no fuel cell module, a standard component in most hydrogen vehicles. No high-pressure storage tanks or refueling pumps are needed, either. Instead, the "hydrogen-on-tap" device contains six stainless steel canisters. Each contains a 113-gram button of an aluminum and gallium alloy. A small amount of water drips onto the buttons, causing a chemical reaction that splits the oxygen and hydrogen contained in the water. The hydrogen releases, and the rest turns into aluminum oxide, a waste product that can be recycled to create more buttons. Back in the garage, the driver can replace spent canisters with news ones to replenish the hydrogen supply. AlGalCo -- short for Aluminum Gallium Company -- has spent 14 years refining the technology, which is based on a process developed by distinguished engineer Jerry Woodall. In 2013, AlGalCo partnered with the Carmel Street Department to build a prototype for one of the city's Ford F-250 trucks. In tests, the red pickup has seen a 15 percent improvement in gas mileage and a 20 percent drop in carbon dioxide emissions.

Read more of this story at Slashdot.

New Slashdot submitter Joe2020 shares a report from the BBC: Firefighters in Corsica, France, are aiming to teach canines how to sniff out coronavirus, as they can other conditions. It's hoped that detection dogs could be used to identify people with the virus at public places like airports. Their trial is one of several experiments being undertaken in countries including the UK and the USA. "Each individual dog can screen up to 250 people per hour," James Logan, head of the London School of Hygiene & Tropical Medicine, told The Washington Post. "We are simultaneously working on a model to scale it up so it can be deployed in other countries at ports of entry, including airports." The dogs are trained using urine and saliva samples collected from patients who tested positive and negative for the disease. "We don't know that this will be the odor of the virus, per se, or the response to the virus, or a combination," Cynthia Otto, director of the Working Dog Center at Penn's School of Veterinary Medicine, told the publication. "The dogs don't care what the odor is ... What they learn is that there's something different about this sample than there is about that sample."

Read more of this story at Slashdot.

Steam will no longer support SteamVR on macOS. The Verge reports: Steam introduced SteamVR for Apple computers way back in the mists of time -- 2017's Worldwide Developers Conference. As The Verge wrote then: "Valve has been working with Apple on this since last summer, which shows a high level of technical and business confidence in Apple's VR efforts." The move was announced in a short post on SteamVR's news page, laid out in a single sentence: "SteamVR has ended macOS support so our team can focus on Windows and Linux." Mac users will still have some access to the feature, however, via legacy builds. One door closes, another will surely open. Right?

Read more of this story at Slashdot.

IT giant accused of paying women less than men doing exact same roles

A lawsuit filed against Oracle on behalf of six women seeking to be paid as much as their male colleagues has been certified as a class action – a legal milestone that will allow thousands of women a chance to have their gender discrimination claims heard.…

The Environmental Protection Agency has rebuffed comments Tesla CEO Elon Musk made concerning what he calls an error during the Model S Long Range's testing process, which the executive says cost the car a 400-mile range estimate. The agency tells Roadshow it conducted the testing properly. CNET reports: During Tesla's Q1 investor call this week -- which also included some colorful language surrounding stay-at-home orders during the coronavirus pandemic -- Musk said the Model S Long Range should boast a 400-mile range estimate, but instead, the EPA gave it a 391-mile estimate. Why? According to the CEO, at some point during the testing process, someone left the keys inside the car and the door open overnight. The Model S entered a "waiting for driver" mode, which depleted 2% of the EV's range, hence the sub-400-mile rating. Musk added that the company plans to retest the Model S with the EPA and is "confident" the test will produce a 400-mile car. The automaker did not return Roadshow's request for comment on the situation, but an EPA spokesperson said in a statement, "We can confirm that EPA tested the vehicle properly, the door was closed, and we are happy to discuss any technical issues with Tesla, as we do routinely with all automakers." It could very well be that Tesla estimates show the Model S Long Range returns a 400-mile range, but for now, the 391-mile estimate sticks with the EPA. To Tesla's credit, that's still the highest range rating of any electric car currently on the market, and just nine miles off the coveted 400-mile mark.

Read more of this story at Slashdot.

Demo-2 mission to send SpaceX capsule, rocket from Florida to the International Space Station this month

NASA today introduced to the world the American astronauts set to ride an American rocket into low-Earth orbit from American soil, a journey that will be the first of its kind since the final Space Shuttle launch in 2011.…

An anonymous reader quotes a report from Ars Technica: Small Internet providers have asked for a government investigation into Frontier Communications' claim that it recently deployed broadband to nearly 17,000 census blocks, saying the expansion seems unlikely given Frontier's bankruptcy and its historical failure to upgrade networks in rural areas. The accuracy of Frontier's claimed expansion matters to other telcos because the Federal Communications Commission is planning to distribute up to $16 billion to ISPs that commit to deploying broadband in census blocks where there isn't already home Internet service with speeds of at least 25Mbps downstream and 3Mbps upstream. An entire census block can be ruled ineligible for the $16 billion distribution under the FCC's Rural Digital Opportunity Fund (RDOF) even if only one or a few homes in the block have access to 25/3Mbps broadband. Frontier's recent FCC filing lists about 17,000 census blocks in which it has deployed 25/3Mbps broadband since June 2019 and tells the FCC that these census blocks should thus be "removed" from the list of blocks where ISPs can get funding. Frontier reported more new broadband deployments than any other provider that submitted filings in the FCC proceeding. The 17,000 blocks are home to an estimated 400,000 Americans. NTCA -- The Rural Broadband Association, which represents about 850 small ISPs, is skeptical of Frontier's reported deployment. "It may be possible that Frontier did precisely what was necessary to meet the standards for reporting significant increased deployment during this eight-month period in the face of years of historical inaction in these areas, admitted shortcomings on interim universal service buildout obligations, and increasing financial struggles," NTCA told the FCC in a filing on Wednesday. "However, such a remarkable achievement warrants validation and verification given the implications. NTCA therefore urges the commission to immediately investigate the claims of coverage made in the Frontier [filing]." The Rural Broadband Assocation went on to say that its members "serve rural areas in the same states as Frontier and, indeed, they frequently field pleas from consumers living in the latter's service area in need of access to robust broadband service. This experience -- and their decades of experience in serving sparsely populated rural areas of the nation more generally -- have caused NTCA members to question whether the filing accurately reflects conditions on the ground changing so quickly in so many places in such a short time."

Read more of this story at Slashdot.

The developers of the open-source QEMU (Quick EMUlator) emulator, which can run programs on various architectures such as ARM and RISC-V, have released version 5.0. Slashdot reader syn3rg writes: Hot on the heels of the 4.0 release (from a major release perspective), the QEMU team has released version 5.0. This version has many changes, including: Live migration support for external processes running on QEMU D-Bus Support for using memory backends for main/"built-in" guest RAMblock: support for compressed backup images via block jobsARM: support for the following architecture features: ARMv8.1 VHE/VMID16/PAN/PMU ARMv8.2 UAO/DCPoP/ATS1E1/TTCNP ARMv8.3 RCPC/CCIDX ARMv8.4 PMU/RCPCARM: support for Cortex-M7 CPUARM: new board support for tacoma-bmc, Netduino Plus 2, and Orangepi PCMIPS: support for GINVT (global TLB invalidation) instructionPowerPC: 'powernv' machine can now emulate KVM hardware acceleration to run KVM guests while in TCG modePowerPC: support for file-backed NVDIMMs for persistent memory emulationRISC-V: experimental support for v0.5 of draft hypervisor extensions390: support for Adapter Interrupt Suppression while running in KVM mode "Not a current user, but I'm happy to see the project advancing," adds syn3rg. For the full list of changes, you can visit the changelog. QEMU 5.0 can downloaded here.

Read more of this story at Slashdot.

A draft bill, first proposed by Sen. Lindsey Graham (R-SC) in January, intends to combat online child exploitation but could introduce significant harm to journalists' ability to protect their sources. TechCrunch reports: Under the Eliminating Abusive and Rampant Neglect of Interactive Technologies (or EARN IT) Act, a government commission would define best practices for how technology companies should combat this type of material. On the surface, EARN IT proposes an impactful approach. A New York Times investigation in September found that "many tech companies failed to adequately police sexual abuse imagery on their platforms." The investigation highlighted features, offered by these companies, that provide "digital hiding places for perpetrators." In reality, the criticized features are exactly the same ones that protect our privacy online. They help us read The Washington Post in private and ensure we only see authentic content created by the journalists. They allow us to communicate with each other. They empower us to express ourselves. And they enable us to connect with journalists so the truth can make the page. This raises the question of whether the bill will primarily protect children or primarily undermine free speech online. It should be pointed out that EARN IT does not try to ban the use of these features. In fact, the bill does not specifically mention them at all. But if we look at how companies would apply the "best practices," it becomes clear that the government is intending to make these features difficult to provide, that the government is looking to discourage companies from offering -- and increasing the use of -- these features. By accepting EARN IT, we will give up our ability -- and our children's future abilities -- to enjoy online, social, connected and private lives. Four of the "best practices" relate to requiring companies to have the ability to "identify" child sexual abuse material. Unfortunately, it's not possible to identify this material without also having the ability to identify any and all other types of material -- like a journalist communicating with a source, an activist sharing a controversial opinion or a doctor trying to raise the alarm about the coronavirus. Nothing prevents the government from later expanding the bill to cover other illegal acts, such as violence or drugs. And what happens when foreign governments want to have a say in what is "legal" and what is not?

Read more of this story at Slashdot.

An anonymous reader quotes a report from NBC News: House lawmakers leading an antitrust investigation into Amazon demanded Friday that CEO Jeff Bezos testify about the company's alleged practice of gleaning financial information from third-party sellers to bolster its own private label business. The House Judiciary Committee threatened to subpoena Bezos if he does not voluntarily agree to testify. The letter comes after a Wall Street Journal investigation found that Amazon employees in the company's private label business had routinely used data from third-party sellers to inform its own product strategy -- a practice that the company has consistently denied to Congress. Amazon's associate general counsel, Nate Sutton, said in a July hearing on the company's practices that "we do not use any seller data to compete with them." He also told Rep. David Cicilline, D-R.I., chair of the House antitrust subcommittee, in the same hearing that "we do not use their individual data when we're making decisions to launch private brands." Amazon has also submitted numerous written responses to the same effect to the committee. "If the reporting in The Wall Street Journal article is accurate, then statements Amazon made to the Committee about the company's business practices appear to be misleading, and possibly criminally false or perjurious," said the House Judiciary Committee in its letter.

Read more of this story at Slashdot.

This week a division of the United Nations announced its new social distancing app designed to help alert people when they get too close to another person during the ongoing coronavirus pandemic. Motherboard has found that the app, called 1point5, is barely functional, and an independent researcher highlighted how the app may be largely ineffective due to how it informs users when they are near any other device which uses Bluetooth, rather than only mobile phones, which a human is presumably carrying in their pocket. From the report: The news highlights the haphazard roll-out of various different apps and technologies that are supposed to help during the pandemic, including those from governments. "1point5 will measure distance to other phones and devices as long as the Bluetooth of those phones is turned on. Those other phones do not need to have the app installed for the app to detect their presence," the press release from the United Nations Technology Innovation Labs (UNTIL), published Wednesday, reads. When users download 1point5, the app is supposed to alert them when another Bluetooth device comes within 1.5 metres of their own phone, or a user can choose to increase the range slightly. The app is then supposed to display a message saying "Please keep your distance" if it detects other nearby devices. But when Motherboard downloaded the app earlier this week before the UN's official announcement, the app didn't even successfully perform this most basic of actions. Motherboard tested the app on two separate Android devices, and held them next to other phones with Bluetooth enabled. The app did not detect any other devices in either test.

Read more of this story at Slashdot.

OpenAI this week released Jukebox, a machine learning framework that generates music -- including rudimentary songs -- as raw audio in a range of genres and musical styles. From a report: Provided with a genre, artist, and lyrics as input, Jukebox outputs a new music sample produced from scratch. The code and model are available on GitHub, along with a tool to explore the generated samples. Jukebox might not be the most practical application of AI and machine learning, but as OpenAI notes, music generation pushes the boundaries of generative models. Synthesizing songs at the audio level is challenging because the sequences are quite long -- a typical 4-minute song at CD quality (44 kHz, 16-bit) has over 10 million timesteps. As a result, learning the high-level semantics of music requires models to deal with very long-range dependencies.

Read more of this story at Slashdot.

Malware maker urges judge to dump lawsuit over WhatsApp phone snooping

Israeli spyware maker NSO Group has rubbished Facebook's claim it can be sued in California because it allegedly uses American IT services and has a business presence in the US.…

Christopher Wray, the FBI director who has been one of the fiercest critics of encryption under the Trump administration, previously worked as a lawyer for WhatsApp, where he defended the practice, according to new court filings. From a report: The documents, which were released late on Wednesday night as part of an unrelated matter, show Wray worked for WhatsApp in 2015 while he was an attorney for the Washington law firm of King & Spalding. While there are sparse details about the precise nature of the work, the filings indicate that Wray strongly defended the need for end-to-end encryption in his previous representation of WhatsApp, the popular messaging application owned by Facebook. Wray's earlier work -- which has not previously been public -- contradicts his current position on encryption, which protects users' communications and other data from being read by outsiders. The Trump administration and major technology companies like Facebook have been at odds over the need to offer customers encryption services, with the White House and law enforcement officials arguing the technology represents a security risk by protecting the communication of terrorists and criminals.

Read more of this story at Slashdot.

LinkedIn is using AI and machine learning to generate screening questions for active job postings. From a report: In a paper [PDF] published this week on the preprint server Arxiv.org, coauthors describe Job2Questions, a model deployed that helps recruiters quickly find applicants by reducing the need for manual screening. This isn't just theoretical research -- Job2Questions was briefly tested across millions of jobs by hiring managers and candidates on LinkedIn's platform. The timing of Job2Questions' deployment is fortuitous. Screening is a necessary evil -- a LinkedIn study found that roughly 70% of manual phone screenings uncover missing basic applicant qualifications. But as the pandemic increasingly impacts traditional hiring processes, companies are adopting alternatives, with some showing a willingness to pilot AI and machine learning tools. Job2Questions is designed to reduce the time recruiters spend asking questions they should already have answers or exposes gaps candidates themselves can fill.

Read more of this story at Slashdot.

NASA and SpaceX's most defining moment of our current space era is coming up at the end of this month, with its Demo-2 mission on May 27. The mission will be the first ever launch for SpaceX with humans on board, and for NASA, it'll mark the first return to U.S.-based astronaut launches since the Shuttle program flew its last flight in 2011. On Friday, representatives from both SpaceX and NASA briefed the media on the mission and the specifics of what it will involve when astronauts Bob Behnken and Doug Hurley board the Crew Dragon for its debut crewed performance. From a report: The first thing to note about this mission is that it's still technically a test, as noted in the "demo" name. This is the capstone demonstration in a series of such missions that will fully human-rate the SpaceX Crew Dragon and Falcon 9 for operational use. As noted during today's press briefings, a big chunk of the actual human rating process occurs during this final mission -- in fact, the majority of the actual final human rating happens on this flight, despite the many years of preparation and live tests to date, including the Demo-1 mission which was essentially a full round-trip flight, just without any astronauts on board. Even though it's technically a demonstration, the stakes couldn't be higher -- SpaceX has a lot to prove here, and it bears the utmost responsibility in terms of keeping Behnken and Hurley safe for the duration of the mission. Which, it turns out, is actually going to be longer than originally planned: NASA says the mission will last anywhere between 30 days and 119 days, depending on a few different factors, the most significant of which being how quickly the agency ends up being able to launch the first operational Commercial Crew mission, Crew-1, which will carry four astronauts, including two from NASA and one from Japan's space agency.

Read more of this story at Slashdot.