« Googlepedia | Main | A manifesto for greener servers »

Falling walls

December 15, 2007

The tipping point approaches.

As Amazon Web Services fleshes out its utility computing package with SimpleDB, Dave Winer reads the economic tea leaves:

Today, when a company raises VC, it's probably because their app has achieved a certain amount of success and to get to the next level of users they need to spend serious money on infrastruture. There's a serious economic and human wall here. You need to buy hardware and find the people who know how to make a database scale. The latter is the hard problem, the people are scarce and the big companies are bidding up the price for their time. Now Amazon is willing to sell you that, to turn this scarce thing into a commodity, at what likely is a very reasonable price. (Haven't had time to analyze this yet, but the other services are.) Key point, the wall is gone, replaced with a ramp. If you coded your database in Amazon to begin with you will never see the wall. As you need more capacity you have to do nothing, other than pay your bill.

Meanwhile, the Sunday New York Times, in a long feature story, lays out the terms of engagement between blood enemies Microsoft and Google as computing shifts to the net's "cloud." First, Google:

So, in Google’s thinking, will 90 percent of computing eventually reside in the cloud? “In our view, yes,” [CEO Eric] Schmidt says. “It’s a 90-10 thing.” Inside the cloud resides “almost everything you do in a company, almost everything a knowledge worker does.”

Then Microsoft:

At Microsoft, Mr. Schmidt’s remarks are fighting words. Traditional software installed on personal computers is where Microsoft makes its living, and its executives see the prospect of 90 percent of computing tasks migrating to the Web-based cloud as a fantasy. “It’s, of course, totally inaccurate compared with where the market is today and where the market is headed,” says Jeff Raikes, president of Microsoft’s business division, which includes the Office products.

No doubt, it's going to take many years before we get any where near Schmidt's 90-10 ratio. Still, in reading Raikes's comments, it's hard not to think of the little Dutch boy with his finger in the dike.


Dave Winer is simply wrong here. SimpleDB is a useful DB store but it's nowhere as powerful as RDMS. If I were investing in a startups I'd be very suspicious of anyone mentioning that SimpleDB is at core of their application. I guess Dave haven't had the time to analyze that either -- too busy blogging.

Posted by: Sergey Schetinin at December 15, 2007 11:36 PM

I don't know how realistic Google's vision of a large majority of computing being done in the "cloud" really is. At this stage of the game, I'm not sure if there's enough of a guarantee of privacy to make me willing to open my company's knowledge up to possible snooping or cataloging by processing my data outside my enterprise. I'm not entirely comfortable with Google maintaining my search history - why would I want to allow them the ability to capture everything my company knows as it flows through cloud?

While the idea of having a shared, distributed, pay-as-you-go set of computing services is appealing, I think it's looked on as too much of a panacea. It reminds me of how outsourcing IT was viewed a decade ago. While outsourcing has worked in some cases, it has definitely not been adopted by anywhere near 90% of the population - and many companies that once outsourced have brought their operations back in-house. The "cloud" will have its successes, but I wouldn't be shocked to see a few high-profile catastrophic failures as well.

Posted by: Dan Turkenkopf at December 16, 2007 11:15 AM

SimpleDB isn't quite simple yet. It makes some things trivial, and other things (that are trivial with RDBMS) very hard. There's still another layer of abstraction that we need before it really removes that wall.

Posted by: EliezerIsrael at December 16, 2007 01:38 PM


I'm surprised to see you reference that Times article -- it was very poorly written and completely misses the real nature of the MSFT vs GOOG relationship (something Steve Gillmor nailed, IMHO).

Office productivity may move into the "cloud", but it's unlikely to be Google's cloud...

Posted by: Sprague Dawley at December 16, 2007 02:55 PM

I'm seeing SimpleDB in terms of Clayton Christiansen's concept of disruptive technology. No, it doesn't have most of the features of a traditional RDBMS. But, for some people, that's not an issue - it's Good Enough. Over time though, I wouldn't be surprised if we saw Amazon grow this thing to be a lot more attractive to developers with more demanding requirements.

There's a definite trade off between features and price/scalability with SimpleDB, but I'm willing to bet it's a compromise a lot of start-ups are willing to make. I know I'd take a serious look at Amazon's web service suite if it meant my application and costs could easily scale at the same pace as demand.

Posted by: Jason Berberich at December 16, 2007 03:33 PM

The desktop vs. in the browser argument does seem to be a curious thing that appears to be approaching the status of a religious schism.

I'm no massive fan of MS, but what the appear to be doing seems to be perfectly sensible.

Google and Microsoft don't seem to doubt that for many users, computing will disappear into the cloud, but what they do differ on is with the front-end.

Google seem to think that we will all access the cloud via the browser. Which will be fine for standard text based apps and simple image editing and of course, you can't beat the convenience factor.

However with this approach you will lose a lot of the power of desktop apps - graphical effect, native desktop integration etc. etc. So Microsoft's approach seems to be trying to combine the best of both worlds.

I'm sure that browser-based GUI APIs will get better and better as the years go by, but it's likely that they will drag behind what native apps can do for a good while longer. So it seems a bizarre step back to suddenly convert all apps to browser based ones and get an experience like Win98/Office 97 era apps.

Incidentally, this also seems to be Apple's approach with the iPhone et al i.e. use Apple's native apps and user experience to access data in the cloud but no-one appears to be berating them re. this strategy as they are with Microsoft...

Posted by: Simon at December 17, 2007 07:50 AM


It's a philosophical issue at heart, and people do tend to get a little extreme about those things.

I can see why a lot of people think Microsoft's approach here is reasonable and balanced, and if this were simply about a market-altering technology I could probably think that, too. But it's not, it's about the philosophy of computing--is the future in your PC, or the cloud?

Put that way, I find it hard to see anything other than the cloud out there. There is a great Steve Jobs quote that I can't quite recall in its entirety, but he was talking about the first time he saw Smalltalk, and how instantly obvious it was at that point that, regardless of the time frame involved, eventually OOP was the way of the future. I don't want to show a failure of imagination, but I think the same of cloud-based computing. Regardless the warts and limitations, the trends involved seem clear and overwhelming.

If you see that, and you think of this as a matter of competing philosophies, then when you look at Microsoft's approach, it's hard not to see it in the same light as IBM's efforts in the PC market in the early eighties. Sure, it was balanced, it looked like a responsible, or maybe even daring, approach at balancing the demands of the broader market at the time with the advent of a disruptive new technology... but we aren't talking about IBM as the juggernaut of the PC industry today, are we? "Balance" was really a fatal hesitation, a failure to abandon a mindset which had made them formidable in pre-PC computing environments, but ultimately proved constraining for them.

Posted by: Scott Wilson at December 18, 2007 12:22 AM

Hi Nick,

I've done a lot of investigation into Amazon's services, and whether they are viable at this stage for a startup. It's sexy stuff on the surface, and investors of small companies always ask about using it, but after a lot of calls to companies who've used the two Amazon services (Ec3 and S2) I've been advised that it's no where near ready for prime time by a web service that needs continuous and steady services. Basically, the companies who were using it said that if it's not your main business (the website) then it's fine, because you can still do business via phone and email, but if it is the main way you interact with customers, if it's your window to the world, then wait a year and see how it goes.

Basically, there aren't good tools yet, even with the one you mention above, to monitor or control and switch as needed, to put something online which can't falter due to "managed hosting" or unmanaged as the case may be.

One company I spoke with, who's had their ecommerce site at Amazon for about 12 months said they had some aspect of their site down 8 days out of the month (that might be the images or the overall site, or the data is so slow to transfer that it's impossible to get).

The bottom line is, you can't rely on it now. Amazon doesn't give SLA's and there aren't tools to swop in and out of Amazon as needed if you use it as a supplementary service, and the monitoring tools are limited but are somewhat available.

I think it's really cool and want to use it, or some other virtualization service, but I won't use it now. It's just not ready.


Posted by: Mary Hodder at December 18, 2007 06:03 PM

Whoops, I meant Ec2 and S3.

Posted by: Mary Hodder at December 18, 2007 06:24 PM

Post a comment

Remember Me?

(you may use HTML tags for style)

carrshot5.jpg Subscribe to Rough Type

Now in paperback:
shallowspbk2.jpg Pulitzer Prize Finalist

"Riveting" -San Francisco Chronicle

"Rewarding" -Financial Times

"Revelatory" -Booklist

Order from Amazon

Visit The Shallows site

The Cloud, demystified: bigswitchcover2thumb.jpg "Future Shock for the web-apps era" -Fast Company

"Ominously prescient" -Kirkus Reviews

"Riveting stuff" -New York Post

Order from Amazon

Visit Big Switch site

Greatest hits

The amorality of Web 2.0

Twitter dot dash

The engine of serendipity

The editor and the crowd

Avatars consume as much electricity as Brazilians

The great unread

The love song of J. Alfred Prufrock's avatar

Flight of the wingless coffin fly

Sharecropping the long tail

The social graft

Steve's devices

MySpace's vacancy

The dingo stole my avatar

Excuse me while I blog

Other writing

Is Google Making Us Stupid?

The ignorance of crowds

The recorded life

The end of corporate computing

IT doesn't matter

The parasitic blogger

The sixth force



The limits of computers: Order from Amazon

Visit book site

Rough Type is:

Written and published by
Nicholas Carr

Designed by

JavaScript must be enabled to display this email address.