Thursday, September 29, 2005

Google's BigTable

Jeff Dean from Google will be giving a talk on October 18 at University of Washington on a new large scale distributed system at Google called BigTable.
BigTable is a system for storing and managing very large amounts of structured data. The system is designed to manage several petabytes of data distributed across thousands of machines, with very high update and read request rates coming from thousands of simultaneous clients.
Sounds very interesting. The talk will be broadcast live on the internet.

See also my previous posts on Google's cool distributed architecture including their file system, cluster, and the distributed data processing tools MapReduce and Sawzall.

By the way, if anyone can find a paper on BigTable, please let me know. I couldn't find one.

Update: The talk was interesting but a little different than I expected.

BigTable stores a distributed, replicated sparse matrix of data. For example, for their crawler, you might have a BigTable matrix with a row "com.cnn.www:WORLD/:http" that contains information about the world news page from CNN. A column for that row might be labeled "content:" and contain the content for that page. Another column might be "language:" and contain "EN" for English. BigTable allows each cell in the matrix to have timestamped data, so a history of changes for the cell can be maintained easily.

It is not, as I was first expecting, a structured distributed database like some Googlized version of MySQL Cluster. That's not what Google needs.

The kinds of data processing tasks that Google has to do everyday require extremely high performance and reliability, but only weak guarantees on data consistency. No database like this exists, so Google had to build their own, BigTable.

Looking at BigTable and Google's other tools, I think Brian Dennis was right when he called them "major force multipliers." Tools like these enable Google to move faster, build more, and learn more than their competitors.

Update: Andrew Hitchcock posted a nice summary of the talk.

Update: The talk is available on Google Video.

Update: Eleven months later, Google has published a paper on Bigtable.

Tuesday, September 27, 2005

Yahoo's Lloyd Braun on personalization

Saul Hansell at the New York Times writes about Yahoo's efforts to deliver interactive TV.

Lloyd Braun, who is leading Yahoo's Media Group, says these efforts will include personalization and recommendations:
One of Yahoo's secret weapons, Mr. Braun says, is that it can personalize information for the interests of each user, such as its My Yahoo page and the song recommendations provided to users of its music service. Mr. Braun is weaving this technology into a video player Yahoo will introduce near the end of the year.

"It will almost be like a television set," Mr. Braun said, except as people watch one program, on the center of the player, other areas will offer additional programming choices, based on their past viewing habits.
Very interesting and, like other personalization, potentially a useful way to surface things that you didn't know about and wouldn't have found on your own.

Yahoo CEO Terry Semel also has a quote in the article where he says that one of the "four pillars" of Yahoo is "personalization technology to help users sort through vast choices to find what interests them."

[via Richard MacManus]

Monday, September 26, 2005

BusinessWeek series on Google

BusinessWeek is running a series of articles on Google, "The Mind of Google's Resident Muse", "Managing Google's Idea Factory" (plus don't miss the related graphic on innovation), and "Google's Search for Simplicity".

The articles focus most of their time on Google's Marissa Mayer, innovation, and managing Google's rapid growth.

First, some excerpts on innovation:
One of the key reasons for Google's success is a belief that good ideas can, and should, come from anywhere. Page and Brin insist that all engineers in the company have one day a week to work on their own pet projects. An ideas mailing list is open to anyone at Google who wants to post a proposal.

What Mayer does is help figure out how to make sure good ideas bubble to the surface and get the attention they need. The task is becoming more complex as Google grows.

At times, [Google's ideas mailing list] more resembles a form of techie Darwinism. Google newcomers who proffer an especially obvious suggestion ("Why don't we search blogs?"), or something off-topic like how to arrange the cafeteria tables, often suffer withering rebukes.

It's all part of a culture not for the faint of heart. Google oozes with what one ex-employee calls "geek machismo." Intellectual sparring can get heated.

What Mayer thinks will be essential for continued innovation is for Google to keep its sense of fearlessness. "I like to launch [products] early and often. That has become my mantra," she says.
This sounds about right. To innovate, you have to let ideas come from everywhere. Let a thousand flowers bloom. However, time is finite. You have to be ruthless, cutting many projects off early, focusing your efforts on the most promising, building on the most successful, iterating and learning as you go.

The articles also say that Google is seeing organizational growth problems not unlike what I talked about over a year ago in "Kill Google, Vol. 1". Marissa Mayer clearly realizes how serious these problems are and appears to be devoting much effort to them:
Some formulas that spawned great ideas with several dozen engineers simply collapse when applied to several hundred, or even thousands of techies. Moreover, indoctrinating a large number of employees with the same sense of opportunity and ethos is a challenge. "As we grow, scale is an issue," she concedes.

Some of the strain is beginning to show .... Without ample guidance or support, some engineers can end up feeling lost in Google's sea of techies. "Things have gotten unwieldy over there," says a former Google engineer.

As Google's ranks swell, convincing techies that they can have just as big of an impact will take work. "It's hard to imagine what it would be like to walk in and start at Google today," Mayer says. "How can you convey that sense of empowerment?"

Google is redoubling its commitment to very small teams. Usually, groups of three engineers will work on even some of the most important projects at the company. By keeping team sizes small -- and teams often share the same office -- Google aims to maintain its nimbleness.
Finally, a tidbit from one of the articles on personalized search at Google:
One of the final groups marches in to discuss a personalized search product. Many pundits describe personalization as the Holy Grail in search. An engine that knows your preferences and interests intimately could tailor the information delivered to improve results. Google has been offering rudimentary personalization for a year, but more is expected in the future.
While MSN and Yahoo talk ([1] [2] [3] [4] [5]) about personalized search, Google is busy delivering it with more to come soon.

Update: Also don't miss this NYT article on Google's internal predictive markets (which appear to be similar to the Iowa Electronic Markets). The teaser for this article on the front page of the NYT business section was "$10 for World Domination". Heh, heh.

Wednesday, September 21, 2005

Findory RSS reader, part II

Findory just finished adding a Bloglines-like feed reader to our personalized news site. It is available on our My Favorites page.

Like most RSS readers, you can import any RSS feed and any OPML list of feeds. You also can easily import your entire list of Bloglines subscriptions by entering your public username or your private username and password.

Like some other RSS readers, we show which feeds have new articles, the new articles in each feed, and which articles you have clicked on.

Unlike many other RSS readers, we're fast, very fast.

Unlike all other RSS readers, Findory shows related articles for every feed, surfacing interesting news articles and weblog posts that you wouldn't have found on your own.

Unlike all other RSS readers, there's no need to laboriously click and skim, click and skim. It's easy to find the most interesting articles. Try clicking "My Top Stories" to see recommended recent articles from your favorite feeds picked based on the articles you read in the past. It's the most important, most recent, most interesting articles surfaced for you from your favorite feeds.

Unlike all other RSS readers, everything you read changes your Findory front page, helping you discover articles from sources you didn't even know existed.

Unlike all other RSS readers, Findory works well for people with hundreds or even thousands of feeds. One of our main test cases was good ol' Robert Scoble and his 1029 feeds. Hope you like it, Robert!

Try it out! If you're a Bloglines user, import your Bloglines subscriptions. If you use another feed reader, import OPML with all your favorite RSS feeds.

If you have any questions or comments, please feel free to write me anytime at I'd enjoy hearing from you.

See also my previous post, "A Findory feed reader".

Tuesday, September 20, 2005

Accelerating Change, AI, and Peter Norvig

Peter Norvig from Google spoke at the Accelerating Change 2005 conference in the "Prospects for AI" panel.

Unfortunately, slides from the talks were not made available, but Scott Lemon provided notes on Peter's talk:
His first point was about Machine Learning ... and joked about the fact that we don't know how to do it. His comments on AI in the Middle included how authors can write trillions of words, systems can detect certain patterns, and intelligent readers can then actually sort through this and find information. He went on to give examples of where apparent intelligence can emerge from larger amounts of data . .. giving examples of the accuracy of Arabic translation based on larger and larger data sets of example translation.
There is tremendous potential in this flood of data, an opportunity to extract knowledge from the noise.

I also liked what Jon Udell said in his brief summary:
Google's "AI in the middle" isn't intelligence per se, but rather a clever mediation between intelligent authors and intelligent readers.
There is wisdom in that crowd. All we need to do is find it.

See also my previous post, "Zen and the art of Amazon recommendations".

Google WiFi and Secure Access

It's being widely reported that Google launched a product called Google Secure Access for encrypting information sent over open WiFi networks.

Many are also speculating that Google soon may be rolling out WiFi nationwide, building on their current experiment in San Francisco.

For details and analysis, see Om Malik, Slashdot, Danny Sullivan, and Nathan Weinberg.

Update: Wired has an article with an amusing blurb about Google's move toward WiFi:
The little search engine that could continues evolving into a Hydra-like monster. It's newest head will chew its way into the wireless internet world, making Google a direct competitor of ISPs and telecom companies.
[Found on Findory]

Monday, September 19, 2005

Yahoo on usability and search

Dan Rose and Jan Pedersen from Yahoo gave a talk at UC Berkeley on "How Search Engines work; Usability and Search" (PDF).

The fourth slide is particularly interesting to me. It says search is an iterative process, a back-and-forth where a person asks for some information, gets something back, refines the query, and so on.

Current search engines don't behave this way. If you search for "personalized news" and then refine the query to "news personalization", the two queries are treated as independent. The information in my history, about what I just found or failed to find, is ignored.

This needs to change. Search engines need to pay attention to what I just did and help me track down what I need.

The rest of the talk is also excellent, a useful survey of what users do when searching. It's clear that users already don't understand the tools available to them, boolean keyword search. This might give pause to those who claim that people just need more powerful search tools, seeing as people don't use or understand the tools they already have.

Thanks to Danny Sullivan for pointing to the SIMS 141 lectures and Professor Marti Hearst for hosting the talks and making the slides available.

I'm looking forward to seeing the next presentations, especially the upcoming talk from Peter Norvig and Sepandar Kamvar from Google.

Update Videos and copies of the slides for most of the SIMS 141 talks are now available. The talks include people such as Susan Dumais, Dan Rose, Jan Pedersen, Marc Najork, and Sergey Brin. Unfortunately, the Sep Kamvar talk was not recorded, darn.

Update: There are also better quality versions of a few of the talks on Google Video.

Netvibes and customizable home pages

A teeny, tiny company called Netvibes launched a nice, drag-and-drop, AJAX heavy, combination feed reader and customizable home page.

It is quite similar to My Google, MSN's, and My AOL, all of which try to be easier versions of the venerable My Yahoo.

Normally with all these giants in the space, a very similar product from a little startup wouldn't be worth talking about. But Netvibes gets a lot of things right. The default page is very clean. It's easy to add feeds, including importing an OPML list, and the UI strikes me as more intuitive and less awkward than the competition.

Unfortunately, the big guys will shamelessly copy the good lessons from this quickly. But, it's some nice work by web developer Florent Fremont and well worth a peek.

All five of these products have the problem Bloglines CEO Mark Fletcher pointed out that they don't track things that you read and don't show you what's new. As Mark says, it's important to know where to focus your attention so you don't waste time.

[via TechCrunch]

Friday, September 16, 2005

MSN and AOL, the kissing behemoths

It's being widely reported that MSN and AOL are in some kind of talks, possibly merger talks, apparently based on an article first published in the New York Post.

I don't know. While I easily could imagine talks between AOL and Microsoft to get AOL to switch to using MSN Search -- AOL web search is currently layered on top of Google -- I have a hard time believing in an merger. Seems to me that mashing these two beasties together for a while will yield nothing but some pretty ugly offspring.

So, yes, on the one hand, MSNAOL (or whatever we might call this Frankenstein monster) would have a huge audience. For many casual and mainstream internet users, AOL and MSN are the web. AOL's tens of millions of users tend to play in AOL's garden. MSN captures all the newbies who don't bother or are too intimidated to change their default browser settings (since Windows is the default operating system, IE is the default browser, and MSN is the default start page).

But the name of the game in web search these days is innovation. AOL and MSN are already on the dysfunctional side of the spectrum in terms of their ability to get cool things done. Merge these two giants and the result will barely be able to control its own chaos. Forget innovation, a merged AOL and MSN would be so clumsy that it couldn't even imitate anymore.

I think Microsoft understands this. They've been working hard on MSN, trying to build a culture of innovation, even experimenting with skunkworks projects like I doubt they'd throw that all away by spilling AOL all over it.

Thursday, September 15, 2005

Bill Gates on attention

In this interview by Jon Udell, Bill Gates had some interesting comments on information overload, attention, and personalization:
RSS things need to need to be able to look at their type in a slightly more predictable way, so you understand which application that things should show up in slightly better, have a little better filtering for these things.

Ultimately the whole problem of notification, of what is it should I be paying attention to next? Is it the e-mail that came in? The phone call? The bid we're supposed to make? That's actually a very deep user interface problem, you know, having all these things understand your context and their priority and who's saying that they think something is urgent. And then you just go to your computer screen and it's ranked for you. You know -- first pay attention to this, then pay attention to this.

That's the holy grail that these technologies are in service of, is that the thing where you always had to go find things, now the system is being a bit smarter for you in terms of now you're not polling the world.
We need help with information overload. We need a smart system that helps us focus on the information that matters. We need personalized information streams where our attention is directed toward the most interesting and important.

See also Sun co-founder Bill Joy's comments on information overload in my previous post, "Making sense of the chaos".

[Found on Findory]

Update: A quote from a May 2006 BusinessWeek article nicely summarizes the problem Bill Gates wants to solve:
[People] are simultaneously overloaded with information that comes at them from all directions, yet challenged to find the key bits of information they need.

Yahoo Instant Search and perfect search

Stephen Hood has the post on the Yahoo Search Blog announcing their new feature Instant Search.

It essentially takes the shortcuts that Yahoo and others provide on a web search and displays them before you even have hit the search button.

It's a cute idea, a nifty little AJAX feature that, like Google Suggest, might provide useful interactivity to normally static web pages.

Unfortunately, the algorithms seem to need a little work. I would think the answers should pop up in any case where Yahoo is sure they have the answer (so it's worth the interruption of a popup) and not in any other cases.

But I found the popup coming up in cases where it wasn't useful (a popup with news on "weather" when I typed "weather" and a popup for news on "capital" when I typed "capital") and not coming up in cases where it should (no popup for "capital washington state").

Despite the flaws, you have to admire the ambition of this move toward perfect search. Ideally, a search engine wouldn't show you a list of results at all, but just the information you need. Clever innovations like this one may be steps toward that goal.

See also comments and reviews from John Battelle, Danny Sullivan, and Charlene Li.

Tuesday, September 13, 2005

Google launches blog search

Chris Sherman reports that Google just launched a weblog search engine. It is available at and

Here's a Google blog search for "findory". Here's a traceback search for all blogs linking to Geeking with Greg.

All the Google operators are supposed to be supported. Doing a traceback is as easy as doing a "link:" query. At some point, I assume you'll even be able to see related blogs using the "related:" operator, but this didn't appear to work in my tests.

The default sort order is by relevance, not by date. Most current blog search engines can only order by date. Relevance seems to be determined by some combination of PageRank and recency, so recent articles from popular blogs seem to bubble to the top. Quite useful, a nice touch. Google's personalized search doesn't appear to be available for blog search yet, but that's probably just a matter of time.

It's super fast. It has support for many languages. It provides feeds so you can subscribe to blog search results. Excellent. I think I just found my new favorite blog search engine.

There have been hints for months that both MSN and Yahoo are about to launch blog search, but their products remain vaporware. It is amusing to see Google quietly beat the others out the door.

This development was widely predicted, but, now that it has become a reality, it certainly doesn't bode well for blog search startups. It was just six weeks ago that Technorati CEO Dave Sifry said in a BusinessWeek interview:
Q. Why can't [Google, Yahoo, and MSN] build ... a blog search engine?

A: Well, good luck. We've been doing it now for almost three years, and it's a lot harder than you think. Doing it on a small scale is not terribly difficult. Doing it to scale becomes pretty hard, and every day the blogosphere is growing by leaps and bounds.
At the time I saw this, I thought Dave's statement was pure hubris. If anyone can handle the scale required for blog search, it'd be Google.

See also my previous post, "Will Technorati die?"

Update: See also reviews and comments by Charlene Li, Anil Dash, Danny Sullivan, Nathan Weinberg, Peter Caputa, and Gary Price.

Update: As Niall Kennedy and others pointed out, one negative of Google Blog Search is that the index only goes back to June 2005. At first, I didn't think this was such a big deal since blog search usually is used for prospective search (current news and information), not retrospective search. This is the idea behind PubSub after all.

But, now that Google replaced the blog search header at the top of this blog with Google Blog Search, I'm finding this June 2005 limit deeply irritating, because I can't use the search at the top of this blog to find my old posts anymore.

Instead, I have to manually do what the search box used to do, I have to do a Google Web Search. For example, notice how this Google Web Search for " perfect search" differs from the equivalent search on Google Blog Search because of the June 2005 limit. Ugh, what a pain.

Update: Stephen Baker at BusinessWeek reviews Google's blog search calling it "lackluster" and saying, "for now, it's just another blog search engine."

But, especially with the other search giants entering soon, I think the KO that Google may have failed to deliver will not be long in coming.

[Found on Findory]

Learning in Microsoft Office 12

Allison Linn at the AP reports that Microsoft Office 12 will "anticipate needs" of users:
With Office 12, due out in the second half of 2006, Microsoft is hoping to entice users with a new system that automatically pops up what it thinks are the most relevant commands based on whether the user appears to be typing a list, editing someone else's work or performing some other job.

The company developed the automated system by tracking ... every keystroke of some Office users.
Could be good, could be another Clippy.

Allison also noted this "is part of an industry-wide trend of trying to personalize technology based on user habits."

Update: In the comments to this post, Gerald Rousselle points out that ex-Amazon Personalization Director Ron Kohavi may be involved in building these personalization features for Office 12.

Update: Jensen Harris, a PM on the Office 12 team, says the AP article got it wrong:
[Myth:] The New UI Tries To Automatically Guess What I Want To Do Next

There was a wire report that was picked up by a gazillion news outlets ... The title of the wire article was "Microsoft: Office 12 to Anticipate Needs".

The reality of the situation is actually exactly the opposite. We looked into all kinds of different interaction design models several years ago when we started this process. Some of them explored the notion of trying to have an even more automatic "auto-customizing" UI that was constantly re-optimizing itself based on usage. The result was as you might guess: unpredictable, unreliable, and meddlesome.

One of the key tenets [of our design doc] was: "No Automatic UI. Prefer predictable, consistent, and human-designed over clever and auto-optimized."
Thanks, Harvey Motulsky, for pointing me at Jensen's post.

Snooping by listening to the keyboard

This is really clever. Ed Felten summarizes a paper (PDF) with a nifty security attack using the fact that keyboards make slightly different sounds when different keys are pressed.

From the paper:
We built a prototype that can bootstrap the recognizer from about 10 minutes of English text typing ... For English text, the language constraints can be applied resulting in a 90-96% accuracy rate for characters and a 75-90% accuracy rate for words.
There's some examples of the recovered text in the paper.

Remarkable and a little scary that, using nothing but a recording of typing and knowledge of patterns in the English language, someone can recover most of what was typed.

[via Bruce Schneier]

Monday, September 12, 2005

Small teams and keeping it simple

I caught some parts of Apple guru Andy Hertzfeld's interview on Robert Cringley's NerdTV. Fun stuff.

Andrej Gregov watched it too and noted a great quote by Andy that I missed:
[A team] should be as small as possible, but no smaller.
Good advice. With larger teams comes additional complexity and overhead. Smaller teams tend to be quicker and more efficient.

As Andrej points out, Andy's quote seems like a variation on Albert Einstein's advice:
Everything should be made as simple as possible, but not simpler.
Which nowadays often is referred to as the KISS principle -- keep it simple, stupid.

Going back a little further, this all may be a variation on Occam's Razor, which is often quoted as
The simplest explanation is the best.
but originally may have been something more like
Entities are not to be multiplied beyond necessity.
Little Web 2.0 startups like Findory live and breathe by these principles. They have small teams, running lean, quick, and efficient. They have simple business structures, often self-funded or minimally funded. They have simple architectures, designed for replication, scaling, and performance. They have clean and easily understood products, tightly targeted at an unexplored niche in the market.

Building something entirely new is complicated enough without throwing additional chaos into the mix. Keep it small, keep it simple.

Update: Mena Trott, co-founder and President of Six Apart, argues the opposite in her post, "In Defense of Big". Her argument seems to boil down to saying that it's less difficult and more pleasant to be working in a big well-funded company. I think that's right, though I'd say the challenge, agility, and opportunity you have when you are small make up for the discomfort.

eBay buys Skype

For $4.1B. That's one costly deal for eBay, an inflated purchase price to acquire a company that has little to do with e-commerce.

[Found on Findory]

Update: Om Malik says:
I don't think, Skype will be the white knight for eBay -- which has been facing executive exodus and general loss of market momentum. I mean, if you can put up a website, and buy Google adwords, you are taking eBay out of the equation. Will Skype help prevent that shift? Not sure, it will.

This is proof the company, which is floundering a bit, is looking for new markets.
See also my previous post, "Google and eBay pursue sellers".

Update: Seven months later, Om Malik says that Skype's revenues are not growing as fast as eBay projected and that "I just can't seem to get the hefty $2.6 billion tag out of my mind."

Update: Two years later, Rachel Konrad at the AP reports that "eBay takes $1.43B charge for Skype" and "$900 million will be ... impairment, essentially acknowledg[ing] that San Jose-based eBay, one of the world's largest e-commerce companies, drastically overvalued the $2.6 billion Skype acquisition."

Friday, September 09, 2005

Larry Tesler and Yahoo's front door

BusinessWeek has an interview with Larry Tesler, Yahoo's new VP of "User Experience and Design".

I like Larry's thoughts on incremental development, learning what works and what doesn't:
We try lots of things. We're trying those small and large changes constantly to all of the pages on our Web site. But before we put them out there, we have to make sure they're actually improvements.
Larry also makes it clear that Yahoo will be doing more personalized content and navigation on their home page:
We've been successful in the last year with adding a little more personalization: If we know where you are, we will give you the local weather. If you frequently visit "In the News," we will put that icon at the top of the page.

We are looking at more ways to personalize the experience. Also, we're always looking at ways of simplifying the navigation. We have taken things off [the home page], and I hope we will find more things to take off, but until we've proven that it improves the user experience, we won't do it.
Over a year ago, I wrote that Yahoo's home page "cries out for personalized navigation" and proposed a some ideas for a redesign that would "emphasize what is most likely to be useful to me." It would be great to see Larry push Yahoo down that path.

I know Larry from his stint at, but many people know Larry from his many years at Apple. Between Apple and Amazon, Larry co-founded a educational software company called StageCast that developed a pretty nifty visual programming language for children.

Thursday, September 08, 2005

Scoble, Bill Gates, and geeks

I found myself enjoying this Robert Scoble interview with Bill Gates.

Scoble's questions struck me as somewhat half-assed, but Bill Gates talked about what he wanted to talk about and came across as deeply geeky. Bill displayed raw enthusiasm about the opportunities ahead of us in computer software.

An excerpt:
The dream of what a PC should be -- we're not there yet ... It's way too hard ... Software has not solved the tough problems ...

Machine learning, natural interface ... We're just touching the surface of it ... Software geeks are the people who get to make those breakthroughs.
Some of this was in response to a question about why, despite being a billionaire, Bill is still working. Bill is working because he loves to solve tough problems.

Some people just don't get this about geeks. We love to solve hard problems. We love the challenge. We love making things that are cool and help people. We build because we love to build.

It's a powerful motivator. It's why Vint Cerf and Louis Monier went to Google. It's why people like me do crazy startups like Findory. We geeks love to build.

After all these years and as huge as Microsoft has become, I was pleasantly surprised to see such geeky passion live on in Bill Gates.

[via Todd Bishop]

Friday, September 02, 2005

Search war or mob war?

Not sure how he got it, but John Battelle posted an excerpt from Mark Lucovsky's testimony in the Kai-Fu Lee case. It has Microsoft CEO Steve Ballmer's reaction when Mark Lucovsky told Steve that he was leaving Microsoft for Google:
Mr. Ballmer said: "Just tell me it's not Google." I told him it was Google.

At that point, Mr. Ballmer picked up a chair and threw it across the room hitting a table in his office. Mr. Ballmer then said: "Fucking Eric Schmidt is a fucking pussy. I'm going to fucking bury that guy, I have done it before, and I will do it again. I'm going to fucking kill Google."
Now, why does this all sound so familiar?
Al Capone: I want this guy dead! I want his family dead! I want his house burned to the ground! I want to go there in the middle of the night and piss on his ashes!

Malone: You wanna know how you do it? Here's how, they pull a knife, you pull a gun. He sends one of yours to the hospital, you send one of his to the morgue. That's the Chicago way, and that's how you get Capone! Now do you want to do that? Are you ready to do that?
When I said the search war was personal, I didn't think it meant this.

Update: This post has been corrected. I originally mistakenly said the statement was from Kai-Fu Lee, not from Mark Lucovsky.

Update: Matt Fletcher at Naive Ameoba made a Flash game called Going Ballmy where you too can throw a chair across the room at a departing Microsoft employee. [via Chris DiBona]