TIME Social Media

Why a Facebook ‘Sympathize’ Button Is a Terrible Idea

Facebook Dislike, Sympathize, Like Button
Brendan Smialowski—AFP/Getty Images A view of Facebook's "Like" button May 10, 2012 in Washington, DC.

It would reduce our empathy to a click

Facebook CEO Mark Zuckerberg isn’t too keen on adding a “Dislike” button to the service, he said in a Q&A event Thursday. But Zuckerberg did float the idea of a new button for “sad moments” when pressing “Like” just doesn’t feel right. That’s why Facebook engineers recently toyed with a Sympathize button, a concept well-received by Facebook’s staff as well as the larger public.

But Facebook hasn’t implemented “Sympathize” yet. It’s still thinking about “the right way” to go about adding such a feature, Zuckerberg said, leaving users hanging. What’s taking so long?

It could be because implementing another Facebook button is a terrible idea — particularly to represent an emotion deeper than “Like.”

Ever since the Like button launched in 2009, the blue thumbs-up icon has become a symbol recognizable by nearly anyone who’s used the Internet. But the Like button’s mega-popularity also resulted in something that wasn’t so stellar: Like Anxiety, which strikes when your posts aren’t getting as many Likes as you think they deserve. While the Like button has made it easier to quickly express emotion on Facebook, Like Anxiety has turned the platform into a popularity contest and insecurity hotbed.

Now try imagining posting about something emotionally crippling — say, the passing of a loved one — and not getting enough “Sympathize” clicks. While most of us can get over when a positive post’s Likes plateau too soon, it would be far harder to move past our sadder missives getting Sympathy-snubbed. Hitting Sympathize is literally the least your friends could do for you in your time of need. If they didn’t click, that would feel pretty awful — you might even start checking which of your friends hit “Sympathize” and which didn’t bother, which wouldn’t be healthy for your friendships.

Facebook’s core mission, as Zuckerberg has put it, is promoting meaningful communications. That goal helps explain why Facebook Messenger was pushed into its own standalone mobile app and why disabling read receipts isn’t an option. Both moves are meant to encourage us to respond to our friends more quickly.

But if Facebook adds a Sympathize button, it would actually make our conversations less meaningful. How? It would override the only way to currently express sympathy on Facebook: Writing a personal comment to a friend, even if it’s only a few words.

The reality, then, is that ‘Sympathize’ is already on Facebook. So is “Dislike.” And “Love.” And “Thanks” — and any other emotion. They’re just not buttons. You have to write those emotions out yourself, and that surely means more than any button ever could.

TIME Innovation

Skype Debuts Instant Translation Feature

A woman communicates with her family abroad by using the Internet telephone system Skype on August 26, 2013 in Stockholm, Sweden.
Jonathan Nackstrand—AFP/Getty Images A woman communicates with her family abroad by using the Internet telephone system Skype on Aug. 26, 2013 in Stockholm, Sweden.

Que? Si.

Skype debuted a hotly anticipated translation program Monday that can translate a conversation between an English and Spanish speaker in real time.

“Skype is now removing another barrier to make it possible for people to communicate irrespective of what language they speak,” announced Skype’s corporate vice president Gurdeep Pall on the company’s official blog.

The program is immediately available to anyone who has a Windows-enabled device and has registered their interest in advance via the Skype Translator sign-in page.

In addition to translating voices and capturing them in text, the program can also instantly translate text messages in more than 40 languages. Skype released a demonstration video of students in the U.S. and Mexico donning headsets and striking up a conversation.

Still, the team noted that the feature is still in “preview mode” and operates by self-learning algorithms that will sort out its translation errors through continual use.

“Our long-term goal for speech translation is to translate as many languages as possible on as many platforms as possible,” Pall wrote.

MONEY cell phones

Americans Spend More on Mobile Service Than the Rest of the World

Person using iPhone at night
Yiu Yu Hoi—Getty Images

We're not only spending more, unlike most countries', our wireless bills aren't going down.

Americans are spending more for wireless services than users in 17 other countries, according to a new report from a UK regulator.

The 2014 International Market Communications Report, published by the UK’s Office of Communications (known as Ofcom), measured the average monthly revenue per mobile connection in 18 different countries across six continents. The report shows that last year, the average mobile customer in the United States paid roughly $47 (£30, in the chart below) per connection, more than cell users in any other nation surveyed.

Only Japanese customers paid close to that amount in 2013, while users in most countries had monthly bills of less than $31 for their service. Mobile users in China, Brazil, Russia, Nigeria, Poland, and India paid less than $16 per month.

Screenshot 2014-12-15 10.30.17

Why are Americans paying so much more for wireless service? One reason is that we tend to use more data than people in other nations. According to data from Cisco, U.S. mobile customers used an average of 1.3 gigabytes of data per month in 2013. In comparison, Europeans used 700 megabytes of data per month—roughly half the American average—and mobile customers in the Middle East and Africa used just 185 megabytes.

But while increased data usage might explain the higher prices overall, it’s less clear why America is one of the few countries where the cost of wireless service isn’t declining. Despite global growth in data use, the United States is one of just five countries Ofcom analyzed (the others are the Netherlands, Sweden, Brazil, Russia, and the U.S.) where revenue per connection isn’t declining. Between 2008 and 2013, the average American customer saw wireless bills rise almost 1% per connection. Countries like France, Spain, and Italy saw cost reductions between 7.5% and 10% during that same time period.

TIME apps

This Is How Uber’s ‘Surge Pricing’ Works

Uber App
The Washington Post/Getty Images UberX driver, Michael Belet, checks the Uber customer app to see where other Uber drivers are working so he can determine where the best place for him to get fares might be, April 7, 2014, in Washington, D.C.

And why it was in effect during Australia's hostage crisis

The backlash against Uber’s surge pricing in the midst of an Australian hostage crisis on Monday was swift and decisive.

Uber riders in Sydney started tweeting about price hikes shortly after an armed assailant burst into a city cafe and took hostages, prompting a massive evacuation of offices and shops in the surrounding area. Fees hiked upwards of four times their normal rate, Mashable reported.

“I understand the way the business works — higher the demand, higher the charge — but four-times at $100 minimum is ridiculous,” one Uber user told Mashable. Almost price gouging at its worst.”

That “price gouging” argument echoed across social media Sunday evening, as it looked to many like Uber was trying to capitalize on a potentially deadly emergency. That conversation added another public relations headache to what’s already been a tough few months for Uber, which is facing lawsuits in several cities and an allegation that a male driver in India raped a female passenger.

Uber’s communication team, realizing what was up, tried to calm the outrage by arguing it was charging riders in Sydney more money to incentivize drivers to show up — which is how Uber has always explained its surge pricing.

“Fares have increased to encourage more drivers to come online & pick up passengers in the area,” read a tweet from Uber’s Sydney account. But many followers found the explanation callous, to put it mildly:

Uber reversed course within an hour, offering full refunds to affected users and free rides to those who were still trying to evacuate the area (Drivers were still getting paid). Even then, Uber added one last defensive note about its pricing policy: “Please note that surge pricing is used to encourage more drivers to come online and pick up passengers from the area,” Uber wrote on its official blog.

That supply-and-demand surge pricing argument has become a common refrain for the company, which has occasionally bumped up against public resistance to its shifting fares. “Seven decades of fixed pricing in car transportation is a lot to unwind in a night,” wrote Uber CEO Travis Kalanick in 2012.

Two years later, questions about surge pricing still linger — such as:

What is Uber’s surge pricing?

Uber’s pricing algorithm automatically detects situations of high demand and low supply and hikes the price in increments, depending on the scale of the shortage. Those higher prices are supposed to make drivers more likely to bite, putting more Uber cars on the road when they’re most needed.

Demand surges have also been monitored by Uber’s human staffers, who have on rare occasions used their discretion to lower prices. After the great Uber Fare Hike of New Year’s Eve in 2012, for instance, Kalanick described the scene at Uber Mission Control. “To our dismay,” he wrote, “the pricing multiplier kept going up.”

“At some point the east coast cities started breaking 6x multipliers—we accepted defeat at that point—the unbending demand breaking our will. We would bring cities down to 3x, only to see conversion go up, supply go down, cars get saturated, and “zeroes” popping everywhere.” Zeroes are Uber’s term for riders who open the app and see no available cars in their area.

Does “surge pricing” get more drivers on the road?

The basic principles of economics would dictate it does — as would Uber’s experience. According to a blog post by Uber board member Bill Gurley, the program has been a success since its inception in early 2012. Uber’s Boston team first tinkered with a price hike on weekend nights around 1 a.m., when drivers tended to clock out just as the city’s public transit system approached closing time, a situation that created lots of demand for Uber cars.

“In just two weeks they had a resounding answer,” Gurley writes. “By offering more money to drivers, they were able to increase on-the-road supply of drivers by 70-80%, and more importantly eliminate two-thirds of the unfulfilled requests.” Economists call this responsiveness to price “elasticity.” Uber’s service does appear to be unusually elastic, given that its fleet of drivers expands and contracts in real time.

Is there a limit to Uber’s surge pricing?

Uber recently announced an algorithm change that sets maximum surge pricing levels during states of emergency in the U.S. When disaster strikes, Uber caps fares at a price that matches the area’s fourth highest price over the preceding two months. Uber has also vowed to donate its 20% commission on rides during emergencies to the American Red Cross.

In non-emergency situations, surge pricing of six to eight times the regular fare have cropped up in moments of extreme shortages. The highest multiple ever recorded was 50 times the regular fare, or $57 per minute, due to an apparent glitch in Uber’s fares in Stockholm, Sweden, Business Insider reports.

$57 per minute? Is that even legal?

Yes, though it can skirt the boundaries of legality under extreme circumstances. At least 34 states and the District of Columbia have passed price gouging laws that forbid businesses from hiking prices in times of extreme emergencies. New York, for instance, can levy penalties of $10,000 for a business that marks up an “unconscionably excessive price” during an “abnormal disruption of the market,” such as Hurricane Sandy.

However, businesses can defend themselves by pointing to higher costs of supplying their service.

Does Uber worry about negative publicity from surge pricing?

As the hostage crisis in Syndey revealed, in extreme times of need, price hikes, rational as they may be, can also unleash a publicity nightmare.


Still, it seems no matter how hard Uber tries to explain its price system, riders’ final fare can still raise hackles. It’s practically become a meme on social media to post an image of a receipt for a three-figure ride, plus an expletive-laced tirade against surge pricing — despite the fact that the Uber app is very clear about surge pricing before users agree to a ride.

Uber board member Bill Gurley pointed out that the company would hazard a far worse form of publicity if it cancelled surge pricing: Chronic shortages of drivers. Better to weather the odd storm, he reasons, than risk a stream of complaints from “tons and tons of unsatisfied customers.”

Uber’s hostage crisis pricing flap may prompt a review of Uber’s pricing policy under extreme circumstances. But for the rest of the time, expect Uber’s surge pricing to stick around.

TIME Retail

Friday Is Amazon’s Free Shipping Christmas Deadline

Inside An Amazon.com Distribution Center On Cyber Monday
Bloomberg—Bloomberg via Getty Images An employee stacks boxes filled with merchandise for shipment at the Amazon.com Inc. distribution center in Phoenix, Arizona, U.S. on Monday, Nov. 26, 2012.

Don't miss these deadlines

The perils of shopping for holiday gifts online or sending packages too close to an impending birthday or holiday are well known. It’s all too easy to click ‘place order’ a day late and leave your Christmas stockings un-stocked because your order didn’t arrive on time.

Amazon, the world’s biggest online retailer, has a huge budget devoted to processing its holiday orders quickly this year. The company recently installed some 15,000 high-tech Kiva robots in its fulfillment centers to whisk items quickly into delivery, and last year, Amazon inked a deal with the U.S. Postal Service to offer Sunday shipping through the season.

There’s no question shipping is squeezing the company’s finances: Amazon spent about $6.64 billion on shipping last year, but only took in about $3.1 billion in shipping payments.

Helpfully, Amazon has a way to make sure you get your online shopping done on time: a simple chart that tells customers when to buy Christmas gifts before it’s too late.

If you live in one of the contiguous 48 states, here are all the last days you can order a present from Amazon if you want to make sure it arrives before wrapping day (December 24).

Free Shipping (Non-Prime) — December 19

Standard Shipping — December 19

Two-Day Shipping — December 22

One-Day Shipping — December 23

Local Express Delivery — December 24

For Alaska, Hawaii and Puerto Rico, check out Amazon’s website.

As far as shipping other gifts to your relatives, the rules are a little more complicated because of all the parcel delivery services in the U.S. But no fear—TIME has compiled a list of handy deadlines for shipping gift packages via UPS, USPS and FedEx to friends and family in time for Christmas Eve arrival. A quick caveat: Unexpected weather and human error can always cause unexpected delays with any shipping service, so the earlier your order your gifts, the safer you’ll be.


The last day to ship with FedEx via the standard ground shipping for pre-Christmas delivery is Wednesday, December 17.

For FedEx Express, the last day is Tuesday, December 23.


Do all your standard package shipping with the government-run postal service by Monday, December 15.

For sending greeting cards and priority mail, the last day is Saturday, December 20.

And for procrastinators out there, Priority Mail Express can be used until December 23.


For standard ground shipping with UPS, send all your packages out by Thursday, December 18.

For UPS 3 Day Select, it’s Friday, December 19.

For UPS 2nd Day Air, you have until Monday, December 22.

For UPS Next Day Air, rely on Tuesday, December 23.

And if things get really urgent and you’re welling to shell out the bucks, UPS Express Critical by December 24 will get the job done.

TIME Security

Sony Asks Media to Stop Covering Hacked Emails

Frederic J. Brown—AFP/Getty Images Pedestrians walk past an exterior wall to Sony Pictures Studios in Los Angeles on Dec. 4, 2014.

Sony says it "will have no choice but to hold [media outlets] responsible for any damage or loss”

Sony Pictures Entertainment demanded Sunday that media outlets stop reporting on the leaked emails and stolen documents obtained by hackers through the recent cyberattack on the studio.

“We are writing to ensure that you are aware that SPE does not consent to your possession, review, copying, dissemination, publication, uploading, downloading or making any use of the stolen information,” attorney David Boies wrote in a Dec. 14 letter that was sent to news outlets including New York Times and Bloomberg, Businessweek reports. If journalists refuse Sony’s request, the company said it “will have no choice but to hold you responsible for any damage or loss.”

MORE: Meet the Sony executive at the center of the worst corporate hack ever

The massive hack on Sony, through which hackers obtained information about salaries, celebrity email addresses and even medical records, was made public in late November. In addition to revealing news about upcoming projects, the emails obtained by the hackers reveal damaging conversations among Sony executives, including some racist exchanges between Sony Pictures co-chair Amy Pascal and producer Scott Rudin that television producer Shonda Rhimes slammed on Twitter.

Oscar-winning screenwriter Aaron Sorkin also spoke out about media coverage of the cyberattack, writing in a New York Times op-ed that reporters writing about information revealed in the leaked documents are “morally treasonous and spectacularly dishonorable.”


TIME apps

France to Ban Uber’s Cheapest Service Next Year

Latest regulatory headache for the ride-sharing app

The French government announced plans Monday to ban Uber’s low-cost service next year, as Paris taxis clogged the capital in a “go slow” or “escargot” protest against the ride-sharing service.

The decision to ban UberPop came after a French court on Friday declined to ban Uber from operating in the country. But Uber’s victory was short-lived.

“Not only is it illegal to offer this service but additionally for the consumer there is a real danger,” French interior ministry spokesperson Pierre-Henry Brandlet told iTELE, questioning drivers’ inadequate insurance. Brandlet said that the ban will begin Jan. 1.

MORE: 5 places where Uber is fighting for its life right now

The decision comes as Uber is facing scrutiny and regulatory pushback around the world. It was banned in Spain, Thailand and parts of India—where an Uber driver was recently accused of raping a passenger — late last week.

Uber did not immediately respond to TIME’s request for comment, although it tweeted Monday morning that demand in France remains strong.

The company, which was fined 100,000 euros in France this October for fraudulent business practices, called some of the country’s attempts to ban the app “discriminatory” last month.

TIME celebrities

Channing Tatum’s Leaked Email Is the Most Adorable Violation of Email Etiquette Ever

Channing Tatum
Evan Agostini—Invision/AP Channing Tatum

Only the Foxcatcher star could get away with abusing the caps lock in this way

The Sony Pictures hack has provided plenty of news about upcoming projects as well as Mean Girls-levels of behind-the-scenes drama, but it’s also revealed how the stars of Hollywood communicate online. In leaked email exchanges that exposed Sony’s plans for a Jump Street-Men in Black crossover film, Jonah Hill wrote, “i think that’s clean and rad and powerful”—capital letters be damned.

But Hill’s Jump Street co-star Channing Tatum appears to have a completely different approach to email. In a conversation with Hill, Sony Pictures co-chair Amy Pascal and other Sony executives, Tatum hit caps lock and didn’t look back as he celebrated the opening weekend of 22 Jump Street, which had the second-best debut for an R-rated comedy in history.

MORE: Meet the Sony executive at the center of the worst corporate hack ever

Below, via Gawker’s Defamer blog, is Tatum’s email, which is a flagrant violation of several of Google exec Eric Schmidt’s 9 rules for email. Then again, Channing Tatum could forward a chain email with instructing you to pass it along to 15 friends by midnight to avoid getting killed by a ghost, and it’d still be charming.

On Jun 14, 2014, at 2:19 PM, “33& out inc C/O FULTON & MEY” <[EMAIL REDACTED]> wrote:



It goes on for way longer than that, but you get the idea.


TIME apps

You Asked: What Is Venmo?

Bloomberg—Bloomberg via Getty Images The Ebay Inc. Venmo application (app) is arranged for a photograph on an Apple Inc. iPhone 5s in Washington, D.C., U.S., on Friday, Aug. 22, 2014.

This online payment service isn’t funny money—it’s the real deal.

For non-social media savvy web users, distinguishing your Twitters from your Facebooks can be difficult.

Twitter’s vice president of data strategy, Chris Moody, has called his network “the most important archives of human thought to have ever existed.” It’s arguable, but if it’s correct, Facebook might be the same, only for actions instead of thought. Facebook is where you pester people to play Candy Crush (please don’t), post pictures of your sandy toes on the beach (yuck), and give everyone updates on how you crushed your latest run (stop showing off).

And recently you’ve starting to see posts like these: “Alex paid Sarah for dinner,” by an app called Venmo. So what’s that all about?

Venmo is a way to send money from person to person through the web. Like PayPal (which owns Venmo), the Internet and mobile device-linked service connects to your major U.S.-based bank account or debit card, so it can pull funds from and make deposits into your checking or savings accounts. Typically used for paying friends back for a bar tab, splitting up the rent check, or settling with the babysitter at the end of the night, it’s designed to take the place of any instance where you would normally use cash or a check to give someone money.

“Venmo’s goal is to make sending money or giving money to people much easier, much simpler,” says Mike Vaughan, Venmo’s general manager. “It doesn’t have to be so hard or complicated when you get money from somebody or have to give money to somebody.”

Of course, ever since society moved away from trading chickens and pelts, cash and coinage has worked great for most purposes. But the explosion of mobile Internet devices and social networks has made cashless payments like Venmo even easier than hitting the ATM. In the last quarter alone, for example, Venmo processed $700 million in payments, a massive number that was up 50% over the previous three months.

I know what you’re thinking: how much of that $700 million does Venmo keep for itself? Nothing, at least when it comes to you paying a friend.

“Venmo was built to be entirely free for that person-to-person experience, and we intend to keep it that way,” says Vaughan. “We’re in effect replacing you handing a $5 bill over to your friend — you don’t pay to do that, and your friend instantly has $5 in his hand.”

But don’t think for a minute that Venmo is a non-profit. As previously referenced, Venmo is owned by PayPal, which also has interests in online-payment processor Braintree. And over the next year of operating under this umbrella, Venmo will start to become a payment method available to users for more than just person-to-person deals.

For instance, says Vaughan, instead of using Venmo to pay a friend when splitting a ride on Uber (which is a Braintree client), users can just pay Uber directly with Venmo.

“The business model then is just the same business model Braintree and PayPal already have created, which is merchants pay to collect for merchant processing,” says Vaughan. “Venmo’s piece of that is really connecting a consumer experience to a merchant that didn’t exist before.”

And coming full circle to the Facebook posts, users can post the bare-bones details of their transactions on Facebook if they want, including who they paid and what for — but Venmo never posts how much money has changed hands. Or, users can elect to keep the transactions private, because not everyone needs to flash their (virtual) cash to friends and followers.

Still, there’s something to be said for the social element of Venmo’s payments. “I can go on in the morning and see what my friends did, where they went last night, because somebody paid somebody back for dinner at a restaurant,” says Vaughan.

Still, posting is just an ancillary feature of this seamless payment platform. “First and foremost, we’re handling people’s money for them, and we have to earn their trust,” says Vaughan. To that end, one of the service’s goals is to keep it simple enough so that users understand exactly what’s going on, not only with their money, but their posts. And there is an option that ensures none of your payments can be seen by anyone else. And as a special request to my Facebook friends reading this: Please check that box.

TIME Social Media

I Harassed My Colleague on Twitter and Here’s What Happened

Bethany Clarke—Getty Images The Twitter logo is displayed on a mobile device.

Taking Twitter's updated reporting tool out for a spin

A few weeks ago, someone called Taylor Swift and me a “dumb b*tch” on Twitter.

On the one hand, I was flattered that he called us a single “dumb b*tch,” because it implies a unified b*tchhood between me and Taylor. On the other hand, nobody calls Taylor Swift a dumb b*tch and gets away with it. So I decided this gentleman’s missive was a good time to test Twitter’s new harassment reporting tool.

The new tool is designed to be much simpler to use than the old reporting method, and it is. The process is streamlined and quick, the menu options are clear, and Twitter responds efficiently to let you know they’ve received your report. What isn’t as clear, however, is what exactly happens to a user who is reported for harassment.

When I reported the Taylor-hater, I quickly got an email from Twitter saying: “This is a confirmation that we’ve received your request. Someone from our team will review it and reply to you shortly.” So far, so good! But what was happening to the honorable wordsmith himself?

Wracked with curiosity, I asked my noble colleague Olivia to help me out. I tweeted at her “U suck,” and asked her to report me for harassment. She did, and she got the same email I did. “They’re investigating it!,” she exclaimed.

Olivia soon received an email saying that my tweet was “currently not violating the Twitter Rules.” So tweeting “U suck” isn’t enough to get you in trouble on Twitter, even if someone reports it as harassment. That seemed like a fair safeguard against haphazard harassment reporting or hyper-sensitive whining.

But it also meant I had to up my game as a harasser.

First, with Olivia’s permission, I called her a “dumb b*tch,” using a decoy account I set up for the purpose. Then — and this I’m not proud of — I said “I’m going to kill you.” Olivia had to warn her mother this was happening so she wouldn’t be alarmed.

The “dumb b*tch” comment provoked no immediate response from Twitter. But the death threat was handled quickly; my (decoy) account was suspended within the day. In other words, Twitter thinks calling someone a “dumb b*tch” is the same as saying “u suck.”

Or, as Taylor puts it: Haters gonna hate, hate, hate hate hate.

I realized I needed a wider variety of harassment options to truly test the response, since there’s a whole lot of ugly in-between insults and death threats. But since my decoy account was already suspended for threatening Olivia’s life, I had to make a whole new account, which I promptly used to harass myself online in increasingly violent and disturbing ways. I can’t embed the tweets (because my account was suspended,) but they said:

“I know you live in Brooklyn and work at Time and your email is [my email address]”

“I hate all Jews including you”

“I hope you die in your sleep”

Charming, right? I reported each of these tweets and others like them to Twitter. That very night, my decoy account was suspended. I could not tweet, my tweets disappeared from my other (real) feed, and I couldn’t change any part of my profile. I couldn’t even embed the tweets (which is why they’re typed out above.)

So from this (very unscientific) experiment, it appears there are two things that will get you quickly suspended from Twitter: death threats and repeated harassment. Some kinds of hate speech against women (“dumb b*tch”) don’t appear to have any consequences. One self-harassment tweet in which I called myself a “C-word” was eventually flagged by Twitter, but by that point, the decoy account had already been suspended.

It seems, then, that Twitter’s new streamlined harassment policy is pretty sane, considering the enormity of the problem its trying to solve — thousands of nasty missives get tweeted every day. It makes sense that Twitter doesn’t suspend people for tweeting things like “u suck,” while it immediately responds to allegations of violence or to repeated harassment. But Twitter hasn’t figured out where to draw the line on non-violent harassment, especially against women, and multiple requests to Twitter Communications to get some explanation were not returned.

A few days after running this experiment, I got an email from Twitter Support about my report on the Taylor Swift tweet. Apparently, calling someone a “dumb b*tch” is “currently not violating the Twitter Rules” — even if that person is Taylor Swift herself.

Your browser is out of date. Please update your browser at http://update.microsoft.com