My favorite thing on the Internets today (aside from Stallman’s tinfoil-hat rant about cloud computing being evil, which I’ll get to later) is this video of the Mythbusters crew researching sobering-up techniques.
When you’re done laughing, think about the first part of that. A member of the media (admittedly, a pretty irreverent one,) showed a roomful of people the illegal content on his hard drive. And they cracked up.
Amazon’s rolling out an extension to its S3 storage offering that will help move content closer to users, reducing WAN latency. “Using a global network of edge locations this new service can deliver popular data stored in Amazon S3 to customers around the globe through local access,” announced Amazon CTO Werner Vogels on his blog. Om beat me to the punch on this one and has a great writeup, too.
The service gives Amazon a much-needed footprint in Asia, but also serves notice to CDN companies that the days of long-term, minimum-rate, negotiated contracts and favored pricing are nearing their end.
[Read more]
Within hours of Chrome’s release, many companies were reporting operational issues. This might seem strange: Chrome is supposed to be leaner, faster, and better. But some of those improvements meant headaches for people running websites — and for those monitoring them. We sat down with Gomez CTO Imad Mouline to look at his company’s experience with the Chrome rollout.
Following its launch, Chrome rocketed to roughly 1% market share practically overnight, according to some sources, and although its use is tailing off a bit, this was a significant enough change in traffic to cause problems. “Small differences under the hood of the browser can lead to big issues in application delivery,” said Mouline. “For example, Chrome has a different connection profile with up to 6 connections per host” which increases TCP session concurrency. “The use of millisecond timing for the Javascript setinterval function also causes issues.”
I don’t want to come across as blindly Googletropic. Google has its wrinkles and warts. But I do admire the way they figure out their mistakes quickly and fix them.
Case in point: The Chrome Terms of Service. Google’s new browser had some pretty onerous terms, namely, that all your data are belong to them. Okay, more specifically, anything you put into the Internets is something they may re-use later.
This makes sense for Google’s mission of indexing the world — today, they only see what they crawl, and what you do on their sites; but much of the Internet happens behind SSL encryption, or in places robots.txt can’t follow. To better advertise to you, they need to know what you’re doing beyond their Googly Appendages.
But Google quickly realized it had overreached, in the same way Facebook did with Beacon. And what I like about Google is that they often correct themselves quickly and completely. They did it when 37Signals compared Appengine demo Huddlechat to its Campfire product. And now, they’ve agreed the TOS is onerous and that they’ll change it, too. Folks like Matt Cutts are a breath of candid fresh air at times when some companies treat online activity like a land grab.
Google’s Chrome browser is pretty sweet. Some of the reasons are simple, some are esoteric; many, people won’t understand.
But the most telling aspect of the release may be the screenshot I’ve attached here. Have a look at the Chrome task manager and the Windows task manager. Which knows more about what I’m doing? Which is on top? Windows is Chrome’s bitch.
Cheekily, Google introduced their new web browser today, aptly named “Chrome” for the sheer lack thereof. Google instead maximizes browser real estate by default, and features the web application first and foremost with a minimalistic set of tabs at the very top of the screen and an “omnibox” serving double duty as an address bar and search window. Nice human factors engineering.
To wax on about the clean user interface would unfairly downplay some more crucial and exciting features under the hood: tabs are actually separate processes with sandboxed priveleges. Finally you can have multiple web applications running concurrently without mucking up each others memory space and fighting for the attention of the browser process. Gone are the days of a single piece of Javascript hanging all 20 tabs you have open and potentially causing you to lose your work. Cross-site scripting just got a whole lot harder.
[Read more]
As Alistair pointed out recently, interest in cloud computing is skyrocketing. All of IT seems to be reframing the industry around the term “cloud computing.” Google’s cloud computing group is a veritable inbox-denial-of-service-machine, and offers more philosophy than practical grist for those of us in the trenches engineering or operating applications in the wol. Reading the threads, I was reminded of this classic exchange between Hamlet and Polonius:
Ham. Do you see yonder cloud that ’s almost in shape of a camel?
Pol. By the mass, and ’t is like a camel, indeed.
Ham. Methinks it is like a weasel.
Pol. It is backed like a weasel.
Ham. Or like a whale?
Pol. Very like a whale.
Companies are scrambling to rebrand their services as cloud based, many of them without substantially changing their offerings. In an intensely speculative industry, overpopulated with caffeine riddled ADD sufferers (it takes one to know one), it is easy to get swept away with the new semantics. But something real is driving the use of the amorphous term, and as this blog has pointed out we have witnessed a seismic transition refocusing the IT industry on a new style of utility computing.
[Read more]
Wagner over at GigaOm talks of a major defection from the world of traditional Massively Multiplayer Online Games (MMOs.) It bears some discussion.
Most MMOs involve software you install (a “fat client”) and a monthly subscription. They’re lucrative — Blizzard’s World of Warcraft is the big one, with an estimated 10 million players worldwide. There are others, like Everquest, Age of Conan, and Eve Online.
These MMOs need to keep adding content to retain users. Warcraft originally let players climb to level 60; then they added the Burning Crusade, which moved the level cap to 70. They’ll soon introduce a third patch which adds more content and takes the cap to 80. Each time, they sell new software and they change the game to make it easier to get to the higher levels, changing game dynamics.
As Wagner points out, the dynamics of the industry are changing. “Casual” MMOs like Runescape have a huge following, and start out free. They run in a browser, and they make their money wherever they can — through item purchases, advertising, and “premium” game models.
If you’re even slightly interested in utility computing — the move towards on-demand, pay-as-you-go processing platforms — then Nick Carr’s The Big Switch is a must-read. You may not agree with everything he says, but his basic thesis is compelling: Just as we went from running our own generators to buying electricity from the power company, so we’re going to move from running our own computers to buying computing from a utility.
Because I spend a lot of my time writing, I’m constantly trying to out-guess the future. And something I’m obsessed with right now is appliances. Not virtual appliances, or network appliances, but simple appliances like pasta makers, bread machines, meat grinders, blenders, and so on.
If you look at the history of the electrical industry, the businesses that became interesting immediately after ubiquitous power was available were those you could plug into it. Generators were boring; but fans, irons, and fridges were really, really cool.
I touched on the topic back in May at Interop (there’s a Slideshare of the deck here on Bitcurrent.) And I think it’s worthy of a lot more consideration because, well, Costco had a sale on Kitchen-Aid mixers.
My wife is an extraordinary cook and an even better baker. And she’s long lusted after a Kitchen-Aid. They’re something of a cult, with a powerful motor, a custom-fit bowl, and dozens of attachments. Most people are happy with a hand-mixer, or a whisk, but there’s an obsessed segment of the market, the Really Serious Home Baker, full of those who simply must have a Kitchen-Aid. So this grey, intimidating, vaguely Cylon-like appliance dominates our countertop.
The Kitchen-Aid is at its core a motor. Its most common use is as a mixer, whisk, or dough hook. But it has attachments that can grind sausage, make ice-cream, roll pasta, shuck peas, and so on. It was conceived in an era where motors were expensive, and attachments were cheap. Here’s a great photo of a precursor to the modern Kitchen-Aid.
Today, motors are cheap. We don’t even think about them. We build them into everything, which is why gift tables at weddings are festooned with single-purpose appliances. And the Kitchen-Aid is the workhorse of near-professionals who demand a 600-watt motor that can tug even the toughest foods into submission.
User interfaces are the modern equivalent of appliances. Until recently, the Internet’s user interface was a desktop computer. Connecting to the Internet was a lot of work for a device: Network signaling, properly rendered graphics, keyboard and mouse, a display with enough resolution, and so on. It required a dedicated machine. The “motor” was expensive, the attachments were cheap. So we put many applications on our PC: Mail, Instant messaging, games, document viewers, file storage, mapping software, videoconferencing, and so on.
But all that has changed. We now have set-top boxes, game consoles, PDAs, cellphones, book readers, SANs — hundreds of devices, all able to access the Internet, all purpose built. That PC in the room is increasingly the jack of all trades, and master of none. The motor is cheap; the attachments matter now.
There are things the PC is still best for: Workstation tasks, like graphic design or software development. But if you want to understand the future of consumer electronics and user interfaces when CPUs are ubiquitous, consider what happened to kitchens when the motor was everywhere.
Self-powered appliances were all about convenience and portability: You don’t have to set up, dismantle, and clean your Kitchen-Aid every time you want to do something, and you can use an immersion blender single-handed over a hot stove-top. In other words, while many cooks crave a Kitchen-Aid, few use it to grind their morning coffee.
We still have to deal with gadget sprawl. Just as everyone has spare hand mixers and blenders secreted away at the back of their kitchen cupboards, so we’re struggling with multiple devices and seeking a way to reduce them. Certainly, high-end PDAs like the Blackberry, iPhone, Windows Mobile devices or the Nokia N95 are tackling this challenge.
It’s also important to remember we’re not just dealing with physical devices, we’re dealing with information. Having multiple blenders isn’t bad–it just wastes space. But having multiple gadgets, each with a part of your digital life on it, is horrible. Which is why synchronization and architectures like Microsoft’s Live Mesh, Google Apps/iGoogle, and Apple’s Mobile Me are so important: It’s not just about decentralizing the physical interface, it’s about decentralizing the information.
When I talk with people about cloud computing and SaaS, I’m always surprised how little mention is made of mobility and ubiquitous computing. To me, these are as big a driver of on-demand platforms like Amazon Web Services or Google App Engine as any of the cost savings or fast development cycles that a cloud can offer.
My wife and I spend a lot of time online. The other day, I lent her my notebook for a few minutes mid-surf, and she quickly went over to Reddit. As it turns out, most of the links I’d opened were the ones she wanted to read anyway. Over at GigaOm, Om’s been reflecting on Facebook for some time now. And this got me thinking.
Surfing is increasingly a social activity. Think of news aggregators as questionnaires: “Which of these stories do you find most interesting?” If we are what we surf, then the people with whom we have the most in common are likely to have similar surfing patterns. This notion alone isn’t particularly revolutionary, and it’s driving innovation in fields like web analytics. But apply it to Facebook Connect, and it opens up a whole new realm of social networking.
Events we'll be attending and covering in the coming months.
Content that has been recently added.
Bitcurrent is part blog, part analyst firm, and part resource site for web operations. We're a loose federation of pundits and entrepreneurs with experience in networking and technology.
