SoylentNews

frojack writes:

According to their Blog SpiderOak ditched Google Analytics for their website a few months back, in favor of their own home grown analytic software running on their own servers. They were also testing Google AdWords, which inserts small ads in pages, for which Spideroak would pay by the click.

One thing they noticed was that their Google AdWords click counts didn't seem to agree with their own analytics and web logs. They were getting billed for clicks that never did arrive at their servers.

Their ads were all in English, but their clicks were from non English speaking countries.

Further when they started comparing the click data from Google to their own analytics they found that half those clicks never arrived at Spideroak servers:

What experiences have other Soylentils had with web advertising that seemed a little questionable?

Original Submission

HughPickens.com writes:

Halfaker, Geiger, Morgan, and Riedl have a new paper on the topic of open collaboration systems about how Wikipedia's reaction to its popularity is causing its decline (pdf).

Open collaboration systems like Wikipedia need to maintain a pool of volunteer contributors in order to remain relevant. Wikipedia was created through a tremendous number of contributions by millions of contributors. However, recent research has shown that the number of active contributors in Wikipedia has been declining steadily for years, and suggests that a sharp decline in the retention of newcomers is the cause. This paper presents data that show that several changes the Wikipedia community made to manage quality and consistency in the face of a massive growth in participation have ironically crippled the very growth they were designed to manage. Specifically, the restrictiveness of the encyclopedia's primary quality control mechanism and the algorithmic tools used to reject contributions are implicated as key causes of decreased newcomer retention. Further, the community's formal mechanisms for norm articulation are shown to have calcified against changes – especially changes proposed by newer editors.

Original Submission

mendax writes:

El Reg is reporting that nature will be giving many of us living northerly climes a fireworks show:

A major eruption from the surface of the Sun could give a spectacular display of the aurora borealis in time for New Year's festivities.

At 1245 UTC (0445 PT) on 28 December our star belched out a coronal mass ejection from the surface in our direction. The National Oceanic and Atmospheric Administration (NOAA) reports that a G3-class solar storm will hit us on 30 and 31 December, and the delightful lighting effects this will cause might be seen as far south as California.

Enjoy!

[Seeing the 'Northern Lights' is one of the top ten things I have ever experienced — if conditions are at all favorable in your area, I cannot too highly recommend watching. I'd expect the 'Southern Lights' will also be visible. Best viewing is generally after midnight, local time. -Ed.]

Original Submission

takyon writes:

The Commonwealth Court of Pennsylvania ruled that a state law (the Older Adults Protective Services Act) banning convicted criminals from getting full-time jobs in nursing homes violates due process rights and is unconstitutional:

The Commonwealth Court in Pennsylvania unanimously ruled Wednesday that a state law that prevents convicted criminals from getting full-time jobs in nursing homes or long-term care facilities is unconstitutional. By a vote of 7 to 0, the court found the law violates the due process rights of otherwise law-abiding people who may have run afoul of the justice system decades earlier. The court also concluded that a lifetime ban on employment for people convicted of crimes is not "substantially related" to the "stated objective" in the Older Adults Protective Services Act—to safeguard the elderly.

The ruling represents a victory for plaintiffs like Tyrone Peake, featured in an NPR story in April. Peake was arrested with a friend in 1981 for trying to steal a car. Peake never got behind the wheel, but he was convicted of attempted theft of an automobile and served three years of probation. The case became a black mark that prevented him from working full-time as a drug counselor for decades. "I've been fired from three jobs," Peake told NPR, "because [of] having a criminal record. And my record is like 32 years old, and I haven't been in trouble since then."

Advocates said the ruling would give fresh momentum to a nationwide movement to allow rehabilitated criminals more access to housing and employment in fast-growing sectors of the economy. Social science researchers say the risk of returning to crime declines as convicts age, and three years ago the Equal Opportunity Employment Commission warned that excluding job candidates because of criminal history can have a disparate racial impact. Aside from Peake, the other plaintiffs had convictions between 15 and 34 years ago for theft, drug possession, writing bad checks and assault and disorderly conduct. Since then, they had each maintained a clean criminal record. But advocates say the law on the books in Pennsylvania treated them the same as murderers and rapists.

More coverage at PennLive. Here's an editorial from a week ago written by a lawyer with Community Legal Services of Philadelphia.

Original Submission

An Anonymous Coward submitted a story published at vox.com:

It was the last semester of my undergraduate degree, and I was broke.

[...] There were plenty of studies to choose from, but one stood out. It was seeking healthy men and women ages 18 to 55 for seven three-night weekend stays in the clinic. Compensation: $5,930.

That was more than enough to cover my rent. It was also enough to buy my ticket to Vancouver, where I was accepted into a graduate program in journalism starting in September. It was even enough to pay for a ticket to see Radiohead in Montreal later that summer, which was my other big priority at the time.

[...] Four weeks into the study, I got a dose of something that wasn't a placebo. A big dose. My heart raced all morning, and the nurse stuck pads on my chest to monitor my heart with telemetry. While everyone else went back to watching TV, I was forced to lie in a bed while the handlers called in a doctor. I didn't think I was going to die, but I was worried that I had damaged my heart.

[...] The good news was that there was no apparent damage to my heart. The bad news was that my reaction to the drug meant I wouldn't be allowed to finish the study and collect the full six grand.

He also regretted to inform me that I wouldn't be allowed to participate in future medical studies.

Would excluding people with adverse reactions significantly increase the probability that future/other drugs pass studies that they shouldn't? If your pool of participants had far more "drug tolerant" people the results might not be representative of the general population or even the target recipients (who might be in poorer health and fitness). Would participants also be tempted to under-report adverse effects in order to stay eligible?

Original Submission

-- OriginalOwner_ writes:

DSLReports notes

Way back in 2005 we profiled the Massachusetts towns of Shutesbury and Leverett, two shining examples of the kinds of U.S. towns that have fallen into broadband connectivity black holes. Large regional providers like Verizon didn't want to upgrade the markets (Boston still hasn't been upgraded to FiOS), and could barely be bothered to keep aging copper in the region fully functional.

A decade later and Leverett last October formally launched LeverettNet, a new network that now delivers up to two gigabit speeds with no usage caps to the town's previously-underserved masses.

The company started by offering locals symmetrical gigabit connections for $65 a month. Starting January 1, locals will now be able two get 2 gigabit connections for $25 a month plus a $50 monthly LLMP operating fee ($75 a month). [* Correction - see update, below.]

Contrast that to Comcast's price tag for two gigabit service: $300 a month with $1000 worth of installation and activation fees. You'll quickly realize why ISPs have turned to protectionist stat laws to ban towns and cities from wiring themselves.

The town also announced this week that it's lowering prices for all of its services (the exact opposite of what usually happens this time of year). The cost of gigabit and phone service is dropping from $44.95 to $39.95 per month, while the price of telephone service is dropping from $29.95 to $24.95 per month (see all prices here). The outfit also announced this week that the project would have a notably lower impact on property taxes than expected.

[Update: We've since confirmed that LeverttNet isn't technically offering 2 Gbps, they're just dropping the price of their 1 Gbps service and associated bundles, and increasing the speed of their POP to 2 Gbps. The last mile connections for these users will remain 1 Gbps. We apologize for any confusion.]

So, what are y'all doing in -your- town?

Original Submission

turgid writes:

Back in the olden days (1960s) there was a British Rail research project to develop a train that could travel at high speed on Birtain's 19th Century railway lines. The project became the Advanced Passenger Train.

The APT employed a tilting mechanism to allow it to go around curves up to 40% faster than conventional trains. It could achieve speeds of 160mph, when not held up by slower traffic. There were even gas turbine-powered prototypes, however in 1981 three electrical trains were built.

Unfortunately, the journalists invited to experience the first Glasgow to London run were plied with drink and reported that the tilting mechanism made them feel sick. Mechanical problems followed, and the trains were withdrawn from service.

They were reintroduced in 1984 but were withdrawn in 1986 for good.

The technology was adopted by other companies in France and Italy, and now Virgin Trains uses the tilting Italian/French Pendolinos on the West Coast Main Line.

[Amtrak in the USA uses tilting technology on the Acela Express line along the Northeast Corridor. -Ed.]

Original Submission

takyon writes:

The BBC reports on a study in the Proceedings of the National Academy of Sciences (PNAS) that examined the sequences of early Irish genomes:

Scientists have sequenced the first ancient human genomes from Ireland, shedding light on the genesis of Celtic populations. The genome is the instruction booklet for building a human, comprising three billion paired DNA "letters". The work shows that early Irish farmers were similar to southern Europeans. Genetic patterns then changed dramatically in the Bronze Age - as newcomers from the eastern periphery of Europe settled in the Atlantic region.

Details of the work, by geneticists from Trinity College Dublin and archaeologists from Queen's University Belfast are published in the journal PNAS [DOI: 10.1073/pnas.1518445113]. Team members sequenced the genomes of a 5,200-year-old female farmer from the Neolithic period and three 4,000-year-old males from the Bronze Age.

Opinion has been divided on whether the great transitions in the British Isles, from a hunting lifestyle to one based on agriculture and later from stone to metal use, were due to local adoption of new ways by indigenous people or attributable to large-scale population movements. The ancient Irish genomes show unequivocal evidence for mass migration in both cases.

[...] "There was a great wave of genome change that swept into [Bronze Age] Europe from above the Black Sea... we now know it washed all the way to the shores of its most westerly island," said geneticist Dan Bradley, from Trinity College Dublin, who led the study. Prof Bradley added: "This degree of genetic change invites the possibility of other associated changes, perhaps even the introduction of language ancestral to western Celtic tongues."

[More after the break.]

takyon writes:

Mark Zuckerberg, CEO of Facebook, has published an editorial in the Times of India defending his "Free Basics" Internet service plan:

Mark Zuckerberg is feeling the force of critics who believe his effort to provide Indians with free access to a limited number of internet services hurts India's democracy and violates net neutrality. In an unusually pugnacious appeal in the mass-circulation Times of India, the Facebook founder forcefully defended introducing his Free Basics service, "a set of basic internet services for education, healthcare, jobs and communication that people can use without paying for data".

Facebook, Mr Zuckerberg says, has already launched the service in partnership with more than 35 mobile operators in more than 30 countries. He says more than 15 million people have already come online because of the service. "The data is clear," he says. "Free Basics is a bridge to the full internet and digital equality."

So - in a tone which many say mocks critics - Mr Zuckerberg asks: "Who could possibly be against this? Surprisingly, over the last year there's been a big debate about this in India."

[...] "Instead of recognising that Free Basics fully respects net neutrality, they claim - falsely - the exact opposite," [said Zuckerberg.]

But prominent tech activists are not convinced. Nikhil Pahwa, a volunteer with savetheinternet.in, says the Facebook boss has not answered a critical question.

"Why has Facebook chosen the current model for Free Basics, which gives users a selection of around 100 sites (including a personal blog and a real estate company homepage), while rejecting the option of giving the poor free access to the open, plural and diverse web?," he wrote in a stinging riposte to Mr Zuckerberg's personal appeal.

Original Submission

https://www.newscientist.com/article/dn28704-sleep-isnt-needed-to-create-long-term-memories-just-time-out/

New memories are fragile. They need to be consolidated before being committed to long-term storage, a process thought to happen while we sleep. But at least some consolidation may occur while we're awake, says Dewar – all you need is a timeout.

The research could have bigger implications for people with amnesia. When Dewar's team conducted a memory experiment with people who had the condition, they saw more striking results. "Most of them can't lead a normal life because they can't remember what they did 10 minutes ago," she says – but all showed huge improvements on the memory test when given a break.

The volunteers were able to recall between 30 and 80 per cent of a list of words when they rested for 9 minutes [pdf]. Without a break, eight of the 12 were unable to remember anything.

Makes me wonder whether trauma counselling really helps or actually makes people remember the bad stuff better ( https://www.myptsd.com/c/threads/experts-urge-against-trauma-counselling-immediately-after-trauma-pre-ptsd.7576/ ). After all many of us memorize stuff by repeating it over and over again and then resting, which is effectively what a lot of such counselling involves. Perhaps people who experience psychologically traumatic stuff and don't want to remember it should tell themselves that this is not important to remember and immediately do a lot of other completely unrelated and more pleasant and interesting stuff to remember - fill the "buffer" or even overflow it... ;)

Original Submission

martyb writes:

Reflections:

Another year has passed. It has been a privilege to work as a volunteer on this site and it is my sincere hope that the staff at SoylentNews have held true to our founding principles of providing a site for people to engage in discussion on topics that interest the community. I know I have learned much from the stories and comments posted here; thanks to you all.

In a nutshell, over the past year we have:

• Journal entries: 717
• Users registered: 1048
• Stories published: 5029
• Stories submitted: 5998
• Comments posted: 152096
• Story hits: 13789650

In addition, there were site upgrades to provide new functionality. Some fine-tuning of comment moderations.

For other data nuggets, check out The Hall of Fame page.

It's all about teamwork, not just milestones. Without all the contributions of every staff member, and the community, this site would not be what it is today. Be aware that this is accomplished entirely with volunteers. So, here's a special shout out to those who keep the site up and functioning, maintain the Wiki, handle the mail system, keep our IRC running, and all the other big and small things that make things work.

Resolutions:

There is a tradition to start the new year with resolutions for the year to come. I'm curious what resolutions my fellow Soylentils have made for the coming year. Lose weight? Learn a new language (programming and/or natural)? Exercise more? Perform more community service? Submit more stories (hint hint ;) to SoylentNews?

I have found that my resolutions have a greater chance of success when I discover others are attempting to make the same changes — it no longer feels like I am alone in the attempt. There is strength in numbers. Also, when I decide to make a change, that's one thing. It's another thing entirely when I tell someone else of my plans — it helps me follow through and not give up. So I encourage you to post your new year's resolutions here.

[Updated to include count of journal entries made and number of times stories were viewed. -Ed.]

Original Submission

"exec" writes:

http://arstechnica.com/business/2015/12/yandex-employee-stole-search-engine-source-code-tried-to-sell-it-for-just-27000-2/

An employee of Russia's Internet giant Yandex, Dmitry Korobov, stole the source code of its search engine and tried to sell it on the black market to fund his own startup, according to a report by the Russian newspaper Kommersant. A Russian court has found Korobov guilty and handed down a suspended sentence of two years in jail.

The Kommersant investigation revealed that Korobov downloaded a piece of software codenamed Arcadia from Yandex's servers, which contained the source code and algorithms of the company's search engine. Later on, he tried to sell it to an electronics retailer called NIX, where a friend of his allegedly worked. Korobov also trawled the darknet in search of potential buyers.

Korobov put a surprisingly low price on the code and algorithms, asking for just $25,000 and 250,000 Russian rubles, or about £19,000 in total. There's no information on Korobov's position within the company, but it appears that he wasn't aware that the data he had in his possession could be worth much more.

[...] Kommersant's sources said that Yandex estimated the value of the code and algorithms at "billions of rubles," or north of £10 million. At court, a Yandex representative said that the software in question "is a key part of our company, it was related directly to Yandex's search engine, which is the main source of the company's income."

-- submitted from IRC

Original Submission

An Anonymous Coward writes:

TechCrunch reports on an announcement posted via Medium by Sidecar CEO Sunil Paul, starting with the announcement, "We will cease ride and delivery operations at 2PM Pacific Time, December 31." From there, an avalanche of self-adulation follows, among which is buried an assertion that the service shutdown announcement is "by no means the end of the journey for the company." Reading between the lines, this appears to indicate that co-founders Sunil Paul and Jahan Khanna are taking their company back into incubator mode. Union Square Ventures partner Fred Wilson previously had blog entries about being "very excited by the potential of shared rides," and that "combining people plus packages is a win/win/win." Re/Code's Mark Bergen adds an anecdote to the mix: "I'd heard that Sidecar was poking around for a sale recently, but did not hear who the prospective buyers were."

Editor's note: Sidecar was a ride-sharing and same-day delivery service founded in San Francisco.

Original Submission

takyon writes:

A new species of bioluminescent lanternshark has been discovered, and it has received a fun name to match its appearance:

A new species of bioluminescent shark has been described from eight specimens collected off the Pacific coast of Central America at depths ranging between 2,700 and 4,700 feet (836 – 1,443 m). A team of marine biologists, led by Dr Victoria Vásquez of Moss Landing Marine Laboratories, chose the scientific name Etmopterus benchleyi for the new species. It also will be known as the Ninja Lanternshark. "The species is named in honor of Peter Benchley, author of Jaws and subsequently an avid shark conservationist," Dr Vásquez and co-authors explained. "His legacy, the Benchley Awards, recognizes outstanding achievements in ocean conservation."

"The suggested common name, the Ninja Lanternshark, refers to the uniform black coloration and reduced photophore complement used as concealment in this species, somewhat reminiscent of the typical outfit and stealthy behavior of a Japanese ninja."

The Washington Post has an article defending the shark's goofy common name.

Etmopterus benchleyi n. sp., a new lanternshark (Squaliformes: Etmopteridae) from the central eastern Pacific Ocean

Original Submission

tonyPick writes:

Spotted at Hackernews, Lifehacker and Good E-Reader is the report that the publisher Springer Science+Business Media has made more than 50,000 digital textbooks available for free download:

Springer has made all digital textbooks more than 10 years old available for free on their website. This equates to thousands of important texts that range in disciplines from cognitive learning to pattern recognition through mathematics.

There has been a ton of conjecture on what free actually means, Springer has cleared the air by updating their terms of service. The company states they are for "You may solely for private, educational, personal, scientific, or research purposes access, browse, view, display, search, download and print the Content.

The book downloads appear to be plain, DRM-free pdf files and the site also works without javascript, offering direct download links. More information on Springer is available on Wikipedia.

Original Submission