Start up: the stuck smart home, McAfee’s hack trick, ICO probes Deepmind deal, Flash the zombie, and more


Yes, Runkeeper tracks your runs. But Norway’s consumer council thinks it tracks more than that. Photo by Gordon on Flickr.

You can now sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.

A selection of 11 links for you. Ain’t that something? I’m charlesarthur on Twitter. Observations and links welcome.

The smart home is stuck • Tech.pinions

Jan Dawson:

»The challenge, then, is the addressable market for most smart home technology is pretty small, composed of innovators and early adopters in the classic technology diffusion curve. As a result, many products are attempting to squeeze every opportunity out of these small markets until they’re maxed out. Nest has been criticized for not innovating more around its original product but I suspect this is the result of a deliberate strategy to saturate many individual product markets rather than focus on ongoing significant improvements in a single market. This helps to explain Nest’s acquisition of Dropcam, its smoke and carbon monoxide detector, and the other products it’s been rumored to be working on. There’s more mileage in opening up new markets than there is in squeezing incremental value out of existing markets already nearing saturation.

I see some people referring to Amazon’s Alexa as a more mainstream smart home or home automation product, and I think that’s actually a red herring. Yes, it can be used to control smart home devices but I suspect (a) only a subset of Alexa devices are used for this purpose and (b) such a focus would limit its appeal to a niche within that smart home early adopter category. I think Alexa’s potential is much broader than that and it’s precisely because it isn’t just a smart home controller. Alexa isn’t extending the smart home market – it’s more mainstream precisely because it’s not limited to that small and limited opportunity.

«

link to this extract


Mobile traffic dominates among the web’s most popular sites • The Atlantic

Adriene Lafrance:

»More than half of Facebook’s roughly 1.7 billion monthly users visit the site exclusively from their smartphones—that’s 894 million mobile-only users each month, up from 581 million such users last year and 341 million mobile-only users in 2014, according to the company’s latest earnings report.

Google confirmed last year that more searches come from mobile devices than computers in 10 countries, including the United States. Over the holiday season, Amazon said more than 60% of shoppers used mobile. And Wikipedia, which recently revamped the way it tracks site traffic, says it’s getting more mobile than desktop visits to its English language site.

In April, Wikipedia had about 361 million unique visits from smartphones and tablets compared with some 229 million from desktops—meaning roughly 61% of traffic to the English-language version of Wikipedia came from mobile devices, according to data provided by a spokeswoman.

«

Didn’t know the Wikipedia stat, but that’s really persuasive.
link to this extract


John McAfee apparently tried to trick reporters into thinking he hacked WhatsApp • Gizmodo

William Turton:

»McAfee has a history of being shifty with the press about his alleged cybersecurity exploits. In March, for instance, during a media tour that included appearances on CNN and RT, McAfee claimed he would be able to hack into the phone of San Bernadino terrorist Syed Farook. McAfee never proved his claims, and later admitted that he was lying in order to garner a “shitload of public attention.” And earlier this year, McAfee hedged on his terrorism-prevention ideals for America during an interview with CNN about his Libertarian candidacy for president, saying that his strategy for preventing homegrown terrorism was “difficult to explain.”

Now, it seems McAfee has tried to trick reporters again, by sending them phones pre-cooked with malware containing a keylogger, and convincing them he somehow cracked the encryption on WhatsApp. According to cybersecurity expert Dan Guido, who was contacted by a reporter trying to verify McAfee’s claims, McAfee planned to send this reporter two Samsung phones in sealed boxes. Then, experts working for McAfee would take the phones out of the boxes in front of the reporters and McAfee would read the messages being sent on WhatsApp over a Skype call.

«

Pointless.
link to this extract


ICO probes Google DeepMind patient data-sharing deal with NHS Hospital Trust • Computer Weekly

Caroline Donnelly:

»The Information Commissioner’s Office, the data protection watchdog, confirmed an investigation into the arrangement is underway, on the back of at least one complaint from the general public.

The deal gives DeepMind access to the healthcare records of 1.6 million patients that pass through three hospitals in North London, which fall under the care of the Royal Free Hospital Trust.

The complaint, seen by Computer Weekly, questions whether DeepMind will be expected to encrypt the patient data it receives when at rest.

“Whilst the information-sharing agreement insists that personally identifiable information – such as name, address, post code, NHS number, date of birth, telephone number, and email addresses, etc – must be encrypted whilst in transit to Google, it does not explicitly prohibit that data being unencrypted at the non-NHS location,” the complaint read.

«

First there’s a deal; then it turns out it’s not directly approved. The complaint is essentially that individuals at Google/Deepmind might access personal data. This is the essential battleground of the coming years: how compatible is tight data regulation with data mining?
link to this extract


Let’s talk about Amazon reviews: how we spot the fakes • The Wirecutter

Lauren Dragan:

»Amazon has a history of trying hard to deal with offenders and shut them down. In fact, in April, Amazon sued another round of companies that are accused of selling fraudulent reviews. But by the time those companies are caught, their clients have already made a bunch of sales, and the fraudulent reviewers will likely pop up again under new names to repeat the process.

(Want to know more? Wirecutter headphones editor Lauren Dragan talks to Marketplace Tech about compensated Amazon reviews and how to tell real crowdsourced opinions from astroturfing.)

You have a few ways to suss out what may be a fake review. The easiest way is to use Fakespot. This site allows you to paste the link to any Amazon product and receive a score regarding the likelihood of fake reviews.

For example, we ran an analysis on some headphones we found during a recent research sweep for our guide about cheap in-ear headphones. You can see from the results below that the headphones’ reviews didn’t score so well.

«

Hadn’t come across Fakespot before; it seems pretty useful.

link to this extract


The real cost of big tech’s accounting games • FT.com

Jonathan Ford:

»How much did LinkedIn make over the past three years? Sounds a simple enough question doesn’t it? But it is also one that is capable of being answered in multiple and very diverse ways.

First, let’s look at the figure the US online networking site wants you to focus on. That’s a mouthful called adjusted earnings before interest, tax, depreciation and amortisation (ebitda), and the total there between 2013 and 2015 came in at a positive $1.7bn.

Sounds pretty hunky dory? Well, now check out the operating profit line for the business — the one calculated according to the generally accepted accounting principles (GAAP) that companies must present but often don’t emphasise. Over the same period, LinkedIn racked up a $67m loss.

What explains the yawning $1.8bn difference between those two figures? It isn’t simply the depreciation and amortisation charges the company took against the value of its assets. Those, while pretty hefty, came to just $791m. No, the biggest single reason for the negative swing was the $1bn cost of the stock LinkedIn stuffed into its employees’ pay packets over those three years.

«

Why does it matter if the company gives stock to employees? As Ford explains, it’s because by doing that

»the firm denies itself the chance to sell those shares or options for value in the market. Failing to recognise that forgone cash effectively understates the cost the company has incurred in employing those individuals.

«

So stock grants are a cost. So they come off the bottom (operating) line. I’m constantly surprised by how many companies’ non-GAAP results are reported as if they were the ones to compare.
link to this extract


Google faces record-breaking fine for web search monopoly abuse • Sunday Telegraph

Christopher Williams:

»Google faces a record-breaking fine for monopoly abuse within weeks, as officials in Brussels put the finishing touches to a seven-year investigation of company’s dominant search engine.

It is understood that the European Commission is aiming to hit Google with a fine in the region of €3bn, a figure that would easily surpass its toughest anti-trust punishment to date, a €1.1bn fine levied on the microchip giant Intel.

Sources close to the situation said officials aimed to make an announcement before the summer break and could make their move as early as next month, although cautioned that Google’s bill for crushing competition online had not been finalised.

The maximum possible is around €6.6bn, or a tenth of Google’s total annual sales.

It will mark a watershed moment in Silicon Valley’s competition battle with Brussels. Google has already been formally charged with unlawfully promoting its own price comparison service in general search results while simultaneously relegating those of smaller rivals, denying them traffic.

«

I’m hearing the same about the timing and intention from my sources; the fine, meanwhile, is indeterminate.
link to this extract


This fitness app tracks you too much, consumer advocates claim • Fortune

David Meyer:

»According to the Norwegian Consumer Council, which has lodged a complaint with the country’s data protection authority, Runkeeper transmits data about its users all the time, not just when the app is in use.

The Norwegian data protection commissioner, Bjørn Erik Thon, confirmed to Fortune that his office has received the complaint and will now look into it.

“Everyone understands that Runkeeper tracks users while they exercise, but to continue to do so after the training session has ended is not okay,” said Finn Myrstad, the consumer council’s technical director.

The data in question includes timestamped location information, as well as Google advertising IDs that can be used to identify the individual.

“Our users’ privacy is of the utmost importance to us, and we take our obligation to comply with data protection laws very seriously,” Runkeeper CEO Jason Jacobs told Fortune. “We are in the process of reviewing the issues raised in the complaint, and we will cooperate with the Norwegian [data protection authority] if it has any questions arising out of the complaint.”

According to the council, Runkeeper’s terms and conditions do not explain how regularly data is transmitted, and users do not give consent to being monitored in this way. The council claims this breaches Norwegian and EU data protection laws.

«

Here’s Runkeeper’s privacy policy. It’s astonishingly vague (though in that respect, probably not so different from other privacy policies). What intrigues me is why the Runkeeper CEO didn’t just say “nah, we don’t collect data after your run.”
link to this extract


Five things you can get in India with a missed call • WSJ

Shefali Anand:

»Want to transfer funds from your account? Give your bank a missed call. Want to hear Bollywood music? Dial a number and hang up.

Making a missed call by calling a number and letting it ring is a popular way of communicating in India because the caller doesn’t have to spend money. Marketing companies, politicians, banks and others now use this practice to reach millions who have cellphones but limited means.

«

Brilliant. Recalls how, in the days when long-distance calls were expensive, kids on their travels would call the operator and ask to set up a reverse-charge call to their parents. Parent’s phone rings: “Alley Okey is calling from Wichita, Kansas. Will you accept the charge?” Parent: “No.” Conversation ends, with parent knowing that the kid is OK and presently in Wichita.
link to this extract


Chinese smartphone market has slowed, but Huawei, Oppo & Vivo have not • Counterpoint Technology

»According to the latest research from Counterpoint’s Market Monitor service, the demand for smartphones in China softened during Q1 2016 (Jan-Mar) as the smartphone shipments were down 2% annually and 13% sequentially.

Commenting on the results, Research Director, Neil Shah, said: “In spite of the Chinese holiday season quarter, the Chinese smartphone market demand reached a standstill. This has led to intense competition between the players as they struggle to take share away from each other. In a market with hundred of brands, growth is now limited to a handful of players with the greatest marketing budgets and headturning designs, and available at competitive price points.

“Only five brands registered healthy growth during the quarter. Oppo, Huawei and Vivo drove the majority of the volume, capturing a combined 40% of the total Chinese smartphone market. Demand for rest of the brands declined, especially Apple after the strong demand for iPhone 6 & 6 Plus in the quarter a year ago, and lacklustre performance from Lenovo, ZTE and Coolpad.”

The Chinese smartphone market saw a lull in the first two months of 2016, however sales for smartphones started to pick up in March, with the largest sales contribution from Huawei, Oppo and Vivo, the new leaders in Chinese domestic market.

«

Other notable points: 98% of phones sold were smartphones (hence Microsoft’s 90% year-on-year drop); the “premium” segment of RMB3000+ ($450+) makes up a fifth of the market, with Apple, Samsung and Vivo dominating.
link to this extract


HTML5 by default: Google’s plan to make Chrome’s Flash click-to-play • Ars Technica UK

Peter Bright:

»In a plan outlined last week, Flash will be disabled by default [in Google Chrome] in the fourth quarter of this year. Embedded Flash content will not run, and JavaScript attempts to detect the plugin will not find it. Whenever Chrome detects that a site is trying to use the plugin, it will ask the user if they want to enable it or not. It will also trap attempts to redirect users to Adobe’s Flash download page and similarly offer to enable the plugin.

«

Great!

»

There will be a few exceptions to this policy, with Google planning to leave Flash enabled by default on the top 10 domains that depend on the plugin. This list includes YouTube, Facebook, Twitch, and Amazon.

«

Crap.

»

Even this reprieve is temporary. The plan is to remove sites from the list whenever possible—Twitch, for example, is switching to HTML5 streaming, so should start to phase out its use of Flash—and after one year the whitelist will be removed entirely. This means that after the fourth quarter 2017, Flash will need to be explicitly enabled on every site that tries to use it.

«

“After the fourth quarter of 2017”, aka 2018. Flash, the desktop web’s malware zombie. (Notice that all those sites somehow muddle through on mobile, which is far bigger, without Flash.)
link to this extract


Errata, corrigenda and ai no corrida: none notified.