One of the main appeals of cloud computing services is the quantized flexibility of their billing schemes — you pay only for the computing resources you need, for the amount of time you need them. That means billing can usually be adjusted on two dimensions: resources required and duration of use.
Now Amazon, ever thinking out of the box, adds a third dimension to the billing flexibility of their cloud computing services: market demand. Amazon Elastic Compute Cloud (more popularly known as Amazon EC2) has launched a new service: Spot Instances. Now you can bid for Amazon’s spare computing resources as they fall within your price range in real time.
Here’s how it works, in a nutshell. Amazon EC2 Spot Instances carry a Spot Price based on current market demand. You can bid on those instances in an open market. Whenever your bid matches or exceeds the Spot Price, your instances start running. Whenever the Spot Price exceeds your bid, your instances are terminated. Persistent Spot Instances will automatically restart once the Spot Price equals or falls below your bid, whereas one-time Spot Instances will simply terminate without restart. Regardless of how high you bid, you never pay more than the actual Spot Price.
As you can imagine, this system results in your rented computing instances running whenever they meet your price, and stopping whenever they don’t. That’s fantastic for controlling costs, but it’s unusable for any application that needs to stay up and running for any fixed continuous amount of time. Unless you want to place ridiculously high bids, don’t expect to run any high-availability systems on EC2 Spot Instances: Web sites, game servers, healthcare systems, battlefield control systems, et cetera.
The silver lining on this ephemeral cloud, however, is that it’s great for any application that doesn’t need to stay up and running for any fixed continuous amount of time: image and video processing, scientific research data processing, financial modeling and analysis, pharmaceutical and life sciences simulations, et cetera.
This eBay-style bidding approach to selling cloud computing resources should prove to be a very interesting experiment. It’ll almost certainly save people a pretty penny on rented computing power, so long as their applications don’t have to be available on demand. If Amazon EC2′s Spot Instances model proves successful, expect other cloud computing providers to come up with their own bidding systems.
When the unlocked, Google-branded Nexus One handset hits the market next year, Google unleashes a two-pronged attack on the Apple iPhone. On one hand, normal people will continue to buy subsidized but locked Android handsets provided by the cellular service carriers. That’s the space the iPhone plays in right now.
On the other hand, hardware geeks will buy expensive but unlocked Google phones provided by the Mountain View powerhouse itself. Thanks to Steve Jobs’ five-year dark bargain with AT&T, that’s a space in which the iPhone simply will not compete.
Thus, Google’s mobile platform allows third parties to maintain walled gardens (carrier-branded handsets with contracts) while maintaining a completely open playground (unlocked handsets). They take on two imporant market segments simultaneously: the common consumers who flock to the walled gardens, and the early adopters in the playground who influence the common consumers. Time will tell if such outflanking will make a dent in the iPhone’s market share.
In the meantime, you have to wonder what the Nexus One bodes for the direction of Google’s upcoming browser-based operating system, the Google Chrome OS, slated for release almost a year after the Nexus One. The potential parallels are certainly intriguing.
Chrome OS will come bundled as an OEM release with very specifically-designed SSD netbooks from various computer manufacturers. Those machines would directly parallel the carrier-locked Android handsets provided by various cellphone manufacturers. If the parallelism holds, could Google release its own hardware for the Chrome OS, the same way it’s now releasing its own hardware for the Android mobile OS? As demonstrated by the Psystar case, Apple is willing to do the latter (provide its own hardware), but not the former (allow branding from third-party manufacturers). The Mountain View geniuses would outflank the Cupertino tyrants once more. Could Google’s two-pronged mobile OS strategy serve as a miniature version (no pun intended) of its future two-pronged desktop OS strategy?
Either way, the current geek hullabaloo over the Nexus One undeniably proves one thing: the existence of a strong demand for Google-branded consumer hardware among the influential blogging-and-twittering digerati. Google can leverage that demand in future consumer hardware projects. We can still expect the bulk of Google Android’s market penetration to come from the carrier-locked handsets, but the geeks will continue to generate buzz about the Google-branded handsets. That’s why T-Mobile and Verizon, who both carry Android handsets, aren’t freaking out right now: the buzz around the expensive Nexus One will help drum up sales for the subsidized G1 and Droid.
Though the actual manufacturing of the Nexus One has been contracted to HTC, there is absolutely no HTC branding on the Nexus One. It’s a Google-branded phone, through and through. High-end hardware manufacturing as word-of-mouth marketing is not a new concept. Though not traditionally a hardware provider, perhaps Google will become the next industry titan to apply that concept — for both Android and Chrome OS.
Remember when LiveJournal was sold to some Russian company nobody’s ever heard of? The once-great blogging pioneer, which was among the first services of its kind to boast such cool features as comment threading and social networking, eventually failed to evolve and fell to the wayside. Now LiveJournal is nothing more than a mass online group therapy session for wrist-slashing emos.
History repeats itself with another pioneer that failed to adapt. If you’ve been into social networking for a while, then you would remember the Great Friending of 2003. That’s the year when Friendster introduced the concept of the online social networking service, when we all rushed to map out our social graphs online for the first time. Heck, Friendster even owns a patent for social relationships, a dubious patent to say the least.
Sadly, Friendster failed to capitalize on its initial success. Instead, like the proverbial hare in a race with the tortoise, the early starter rested on its laurels and failed to follow through on two very important things:
1) Scalablity. Friendster’s scalability problems started with cache replication errors here and there. Then they got worse. Dozens of friends would disappear at a time. Updates would fail to show up, a no-no in the realtime Web. Eventually, it got so bad that the whole site went down for four days without explanation last year. Friendster’s failure to scale gave MySpace the opportunity to rocket ahead in the social networking space.
2) Interoperability. Facebook introduced its application platform. MySpace introduced MySpaceID. Twitter’s rich APIs spawned hundreds of mashups. While all that was happening, Friendster remained… Friendster, the same old social information silo it always was. That’s how Facebook became the world’s top social network, with 350 million users using Facebook Connect 60 million times a month across the Web. That’s why over six thousand sites across the Internet — from huge sites like The New York Times to tiny sites like my own personal blog — now support Facebook Connect, and not Friendster Connect.
Friendster’s spin doctors would have you believe the site still rules Southeast Asia. That’s a lie. It’s the same kind of lie Carol Bartz tells about Yahoo: commercial strength among the unwashed and disengaged masses. I’m currently typing this post from the tropical paradise of the Philippines, surrounded by all its beautiful women — the most beautiful of whom are on Facebook and MySpace and Twitter. Around here, Friendster is considered the online hangout of the lower class. Think about the kind of people you see on MySpace, scaled down to the third world. You’ll get the picture. That’s why your next social network ad campaign to target Southeast Asia will probably be on Facebook, and not on Friendster.
I actually dread the thought of Friendster rolling out a federated identity service, something like Facebook Connect or MySpaceID or Twitter OAuth. Friendster is filled with moronic teenagers who type in mIzSpElT cAmElCaSe. God forbid Friendster Connect spread them across teh Intarwebz.
Even after all these years, Friendster still hasn’t figured out that it needs interoperability, that social information silos just won’t cut it in the age of the mashup, that it takes more than just fancy layouts and teenybopper marketingspeak to succeed in an interoperable medium like the Web. From its cutesy unreadable logo with vague cloud computing allusions to its Twitter-wannabe front page, Friendster’s new interface looks like a parody of Web 2.0 design motifs.
The most recent ad for Friendster even flies in the face of Metcalfe’s Law, actually mocking Facebook’s popularity and trumpeting Friendster’s status as a hopelessly insular, disconnected, geographically isolated information silo. Apparently, Friendster is only for Asian teenagers with no concept of data portability or people outside their little cliques — and they’d like it to stay that way, thank you very much. White/black/Hispanic people who’ve been spoiled on the power and reach of the Facebook platform aren’t welcome at Friendster’s little kiddie clubhouse.
For all these reasons and more, Friendster has met a fate similar to that of LiveJournal years before: sold off to some Malaysian company nobody’s ever heard of, all for the piddling sum of $110 million. Oh yeah, Facebook founder Mark Zuckerberg is just quivering in his boots. After all, Facebook is only worth $15 billion.
Even Zynga, maker of the popular Facebook game Farmville, makes only $250 million per year. Oh yeah, Zynga founder Mark Pincus is just itching to drop Facebook in favor of Friendster. To quote the new Friendster ad, whoop dee doo.
Let the sad fate of Friendster serve as a lesson to founder Jonathan Abrams, and all pioneers who whine about their supposed entitlements while smart competitors leave them in the dust. Nobody cares if you were the first to do social networking, or blogging, or sliced bread, or fire, or the frickin’ wheel. Seniority does not equal relevance. Evolve or perish.
As ever more sophisticated search, subscription, recommendation, and social graphing tools personalize our online experiences with individually-tailored content at the risk of tunnel vision, as the speculative fiction of the Daily Me comes closer and closer to our day-to-day realities, we sometimes lose sight of just how big the Internet really is. Encircling the globe like the World Serpent of ancient Norse mythology, this enormous and untamed beast of fiber and code enables us to create more knowledge in a single year than generations of our ancestors did throughout the last six millenia of recorded history.
Just look at a typical day on the Internet, as illustrated in the following infographic.
Once again, Verizon plays up the hardcore techie appeal of the Google Android-based Motorola Droid. Their latest advertisement contrasts the Droid’s raw mechanistic power with the iPhone’s glitz and glamour. In the words of this particularly edgy ad, the iPhone is a “digitally clueless beauty pageant queen”.
Contrary to Apple’s claims, more than one frustrated developer would probably agree with that assessment. After all, the iPhone app approval process is as fickle and arbitrary as any blonde bimbo. Sure, you’d show off your beauty queen girlfriend in public, and you’d certainly have lots of fun with her in private — but she probably wouldn’t make the most engaging conversationalist, or the most effective executive assistant.
With the open application development offered by the Google Android mobile OS, Verizon is clearly positioning the Motorola Droid not as a buxom blonde beauty queen, but as a fluent and flexible executive assistant. That certainly jives with their thrust to get developers and early adopters using the thing. Geeks love girls with glasses.
It’s days like these that make me really, really happy that so many Google services are so indirectly monetized. OpenDNS, the world’s most popular DNS service, makes money by hijacking 404 errors to show its own ads. Combined with revenue from filtering services for corporate users, OpenDNS made $20,000 per day on just 7 billion daily queries last year. Today, they handle 20 billion daily queries.
Sure, OpenDNS is a great service that offers lots of features, but DNS hijacking is an ethically questionable practice. Error pages full of ads might occasionally be useful, but they’re one step away from hijacking legitimate pages — for censorship, phishing, blocking competition, whatever nefarious purpose.
Enter Google Public DNS, which follows DNS protocols to the letter: absolutely no hijacking, even for error pages. If you misspell stuff in your address bar, or if the webmaster of your favorite site is a moron, then you get an error page and nothing else. If Google isn’t showing ads through Google Public DNS, then what do they gain from offering this service for free? Three things: More →
Now, to demonstrate the gaming capabilities of your new N900 smartphone, you’ve resorted to running classic games on retro console emulators. Never mind the sheer absurdity of using software emulation to sell the native capabilities of a piece of hardware. Sure, emulators are legal, and hardcore mobile gaming aficionados love emulators, but the game ROMs in your promo video are of questionable legality.
Now Nintendo’s thinking of taking you to court for copyright infringement. Considering that you’re the world’s largest cellphone manufacturer, that’s pretty frickin’ embarrassing.
If you thought the Windows 7 house party ad was lame, wait ’til you see this. On November 8, Microsoft was planning to sponsor a special episode of Seth McFarlane’s animated comedy series Family Guy to promote their latest operating system.
You can see where this was going: Family Guy’s humor is very edgy, while Microsoft’s marketing is very corporate. Combining the two would have resulted in a horrifying trainwreck of pure sellout stupidity.
Fortunately, Microsoft pulled out of the deal after executives from Redmond saw an actual taping of the show. Apparently, up until that point, they were blissfully unaware of Family Guy’s edgy humor. Talk about being out of touch.
Now Microsoft has released clips from that hideous abortion through the Windows YouTube channel. As you can see from these clips, Microsoft made the right decision by killing their collaboration with McFarlane. One of the video descriptions even acknowledges the blatant selling-out involved: “When Brian notices Stewie has a new PC, he thinks he smells a sellout.”
Call your neurologist, because the sheer inanity of these totally unfunny sketches will kill your brain. More →
Ah, the Sony PlayStation 3. A Blu-Ray player, an online video player, a Linux box — it’s used as anything but what it’s supposed to be: a game console. Now the US Air Force is buying 2,200 PS3s to build a supercomputer.
No, I’m not kidding.
The U.S. Air Force is looking to buy 2,200 Sony PlayStation 3 game consoles to build out a research supercomputer, according to an document posted on the federal government’s procurement Web site.
The PlayStation 3s will be used at the Air Force Research Laboratory’s information directorate in Rome, N.Y., where they will be added to an existing cluster of 336 PlayStation 3s being used to conduct supercomputing research.
The Air Force will use the system to “to determine the best fit for implementation of various applications,” including commercial and internally developed software specific to the PS3′s Cell Broadband Engine processor architecture. The research will help the Air Force decide where Cell Broadband Engine processor-derived hardware and software could be used in military systems.
You might think that Sony’s the big winner in this deal, and that the Air Force is stupid for buying thousands of overpriced game consoles. You’d be wrong, on both counts. Sony sells PS3s at a loss, hoping to make their money in games and accessories. That means the Air Force is getting Cell hardware for a fraction of the cost. More →
Entertainment industry execs, take note: this is how an old classic can successfully make the transition into new media. When The Muppet Show was cancelled in 1981, a lot of us felt like we lost our old gang of childhood friends. Fortunately, the Muppets just launched their own YouTube channel, and they’re hitting the medium of online video with the force of Gonzo shooting out of a cannon.
Kermit and crew clearly did their homework. With punchy new original content perfectly tailored to the viral, shortform, fourth-wall-breaking nature of online video, the Muppets know exactly what they’re doing on YouTube — despite Statler’s and Waldorf’s constant assertions to the contrary. Apparently, old frogs can learn new tricks.