Thursday, May 31, 2012

Goal-line Technology for soccer

Here's a reason to root for a high-scoring England-Belgium friendly this weekend: With Goal-Line Tech, Soccer Tries Kicking Its Addiction to Human Error, since if the game should end 0-0, it's not clear what we will learn about the Hawk-Eye system.

The camera system developed by the British firm Hawk-Eye Innovations could revolutionize soccer by tracking and triangulating the ball’s position, leaving no doubt when it has crossed the goal line. Although the tech will be deployed before a crowd of 90,000 at Wembley Stadium, game officials won’t use it to settle disputes. The data will be examined only by scientists during the final test before soccer’s governing body decides on July 5 whether to adopt goal-line technology.

Of the two competing systems, one uses cameras and image processing, while the other uses electromagnetism:

The two companies declined to discuss their technology in detail pending the IFAB’s decision. But Hawk-Eye, a system used widely in tennis and cricket, places 14 cameras around the pitch to triangulate the ball’s location. When the entire ball crosses the plane of the goal, a radio signal confirming the goal is sent to the ref’s watch. The downside is at least 25 percent of the ball must be visible so the cameras can make the ruling. GoalRef, a joint German-Danish project, uses a microchip in the ball and a magnetic field around the goals. Once the entire ball has crossed the line, the change in the magnetic field signals a goal.

A Very Short Review of The Hangman's Daughter

By way of the quite-nice Kindle Owners Lending Library, I recently read Oliver Potzsch's The Hangman's Daughter.

At least part of my interest in the book was that, at the time, my parents were traveling in South Germany, and Potzsch's book is set in the same area (although hundreds of years earlier in time), so it was fun to read about "Augsberg this" and "Ulm that" and "Ingolstadt the other thing".

But Potzsch's book has a lot to recommend it, even if you're not trying to find a fun murder mystery set in Bavaria.

For one thing, the central character is intriguing: Jacob Kuisl is the town hangman for the Bavarian town of Schongau.

In this time, in this part of the world, the town hangman is one part policeman, one part jailor, one part pharmacist, and one part garbageman. The townspeople routinely seek Kuisl out for all sorts of useful (but unpleasant) services, while simultaneously shunning him as low-caste.

In Potzsch's telling, Kuisl is a fascinating individual: self-aware, learned, cynical, and yet morally strong.

When the town is visited by a series of abductions and murders, the townspeople cry "Witchcraft!" and thirst for revenge, and Kuisl, together with a small set of colleagues, must conduct his own research to solve the actual crime, while trying to endure the narrow-minded prejudices of the townsfolk.

The book is rich in atmosphere, nicely paced, and great fun to read, and I really liked learning about this side of medieval life that is rarely described in this manner.

Looking for a summer vacation read? Give The Hangman's Daughter a try.

Wednesday, May 30, 2012

Anand-Gelfand: Anand retains his title

Today in the World Chess Championship, the format was 4 rapid games. Three of the games were draws, but Anand won the second game (with White).

I believe that this concludes the championship match, and Anand is World Champion of chess for another two years.

Tuesday, May 29, 2012

More data generally does in fact help

For example, when the mystery of the Pioneer Anomaly was resolved, it was due to having more data: The Mystery of the Pioneer Anomaly Solved at Last

But for their new analysis, Turyshev et. al. compiled a lot more data than had ever been analyzed before, spanning a much longer period of the Pioneers' flight times. They studied 23 years of data from Pioneer 10 instead of just 11, and 11 years of data from Pioneer 11 instead of 3. As explained in their new paper, the more complete data sets reveal that the spacecraft's anomalous acceleration did indeed seem to decrease with time. In short, the undying force had been dying after all, just like the decaying plutonium.

50 yaks to carry the gear back down

Here's a nice report from the MountainTrip team on their successful summit of Mount Everest: Back to Base Camp!.

Less than two weeks til the 2012 Euros!

The competition starts June 8th, I believe.

Let's hope that fears like these turn out to be unfounded.

An exposé by the BBC's Panorama program, broadcast in the United Kingdom on Monday night and set to be shown again in multiple countries, detailed the extent of race-related soccer hooliganism in Poland and the Ukraine, where the Euros will be jointly staged starting June 8.

Monday, May 28, 2012

Flame on!

I'm puzzled by the use of the word "discovered" in this paragraph from Kaspersky's breathless press release announcing their work on the new Flame malware:

The independent research was initiated by ITU and Kaspersky Lab after a series of incidents with another, still unknown, destructive malware program – codenamed Wiper – which deleted data on a number of computers in the Western Asia region. This particular malware is yet to be discovered, but during the analysis of these incidents, Kaspersky Lab’s experts, in coordination with ITU, came across a new type of malware, now known as Flame.

What does it mean to have "incidents" with an "unknown, destructive malware" which is "yet to be discovered"? What is discovery if it is not the observation of a malware-caused incident?

The Kaspersky "Flame FAQ" is fascinating:

there are internally used local databases with nested SQL queries, multiple methods of encryption, various compression algorithms, usage of Windows Management Instrumentation scripting, batch scripting and more.

It will be very interesting to see if any other independent researchers can confirm these findings, as they seem rather unusual on the face of it. The only other reports right now appear to be from the Budapest University of Technology at their web site: http://www.crysys.hu/

Anand-Gelfand game 12: draw

Anand had White and opened 1. e4. He varied from game 10 on the fifth move. The queens came off the board on move 13, and the game was drawn on move 22.

And so, the classical-format portion of the match has completed, with the score tied 6-6.

The format now changes, and they will play games at a shorter time format. I'm not exactly sure what that new format is, and how the games will be scheduled. Here is how the Times of India describes it:

In the rapid chess, Anand will have to score 2.5 points in the four games scheduled on Wednesday. There will be 25 minutes to each player in this contest with a 10-second addition after every move is played.

If the scores are still tied after these four games, there will be two more games with blitz chess rules. Should the tie persist, there are five such blitz matches to be played.

If the deadlock still continues, there will be an Armageddon game with five minutes to white and four to black and black wins the title in case of a draw also.

25 minutes per side with a 10-second increment is much faster than the classical time, but each game could still take an hour to complete.

Anand's skill at the faster time controls is legendary, but Gelfand's play in the match has been superb.

What will Wednesday reveal?

Saturday, May 26, 2012

Things I'm reading this weekend

In between the barbecuing and the gardening and the walks with the dog, a few things to read:

  • Michael Wolff: The Facebook Fallacy
    The subtext—an overt subtext—of the popular account of Facebook is that the network has a proprietary claim and special insight into social behavior. For enterprises and advertising agencies, it is therefore the bridge to new modes of human connection.

    Expressed so baldly, this account is hardly different from what was claimed for the most aggressively boosted companies during the dot-com boom. But there is, in fact, one company that created and harnessed a transformation in behavior and business: Google. Facebook could be, or in many people's eyes should be, something similar. Lost in such analysis is the failure to describe the application that will drive revenues.

    And don't miss Doc Searls's take on Wolff's article: After Facebook fails

    Three problems here:

    1. By its nature advertising — especially “brand” advertising — is not personal.
    2. Making advertising personal changes it into something else that is often less welcome.
    3. There are better ways to get to achieve Michael’s objective — ways that start on the buyer’s side, rather than the seller’s.
  • Chris Clark's essay on an unexpected side-effect of performance improvements in the Windows 8 boot process: Designing for PCs that boot faster than ever before
    However, the hardware and software improvements in Windows 8 have collapsed the slice of time that remains for Windows to read and respond to the F8 keystroke. We have SSD-based UEFI systems where the “F8 window” is always less than 200 milliseconds. No matter how fast your fingers are, there is no way to reliably catch a 200 millisecond event. So you tap. I remember walking the halls and hearing people frantically trying to catch the F8 window – “tap-tap-tap-tap-tap-tap-tap” – only to watch them reboot several times until they managed to finally get a tap inside the F8 window. We did an informal study and determined that top performers could, at best, sustain repeated tapping at about a 250ms frequency. Even in this best case, catching a 200 millisecond window still depends somewhat on randomness. And even if you eventually manage to catch this short window of time, you still have to contend with sore fingers, wasted time, and just how ridiculous people look when they are frantically jamming on their keyboard.
  • David Lowery's essay on the economics of the music industry in the age of the Internet: Meet The New Boss, Worse Than The Old Boss?
    My only explanation is that there is just something fundamentally wrong with how many in the tech industry look at the world. They are deluded somehow. Freaks.

    Taking no risk and paying nothing to the content creators is built into the collective psyche of the Tech industry. They do not value content. They only see THEIR services as valuable. They are the Masters of the Universe. They bring all that is good. Content magically appears on their blessed networks.

    I’m using this language for good reason. There is a quasi religious tone to many tech convention speeches and press releases. What other industry constantly professes utopian visions for all humanity? What other industry would dare proclaim they were liberating artists? Students? Workers? What other industry thinks they are mystical shaman “Let’s send our magic objects, our laptops to poor children in third world countries”. What other industry genuinely believes they (and only they) possess the lapis philosophorum? They have even created their own God. A Superhuman intelligence that they (naturally) have created. The singularity. Their egos know no bounds.

    Not only is the New Boss worse than the Old Boss. The New Boss creeps me out.

  • Google's description of how an email is actually processed: The Story of Send
    Once your message leaves your Internet Service Provider (ISP), it enters an Internet backbone router. Here’s where Google picks up your message and guides it to the closest Google data center. To provide the best possible user experience, we try to pick up your requests as early as possible from the local ISP, so we’ve built an extensive Internet backbone across the U.S.

    And again, Doc Searls has a nice follow-up: The Real Story of Send

    The main intended message of The Story of Send is a green one: Google saves energy. A secondary message is that Google is a big nice company that treats your mail well and has good security practices. But the main unintended message — or at least the one that comes across — is that email is a big complicated business, and you need big complicated companies to do it right. It also ignores the real story, which is about a handful of simple protocols.
  • Jorge Lucangeli Obes and Justin Schuh's nice essay about the details of one of the Chrome security holes that was recently found and patched: A Tale of Two Pwnies (Part 1)
    The next thing Pinkie needed was a target that met two criteria: it had to be positioned within range of his overwrite, and the first eight bytes needed to be something worth changing. For this, he used the GPU buckets, which are another IPC primitive exposed from the GPU process to the Native Client process. The buckets are implemented as a tree structure, with the first eight bytes containing pointers to other nodes in the tree. By overwriting the first eight bytes of a bucket, Pinkie was able to point it to a fake tree structure he created in one of his transfer buffers. Using that fake tree, Pinkie could read and write arbitrary addresses in the GPU process. Combined with some predictable addresses in Windows, this allowed him to build a ROP chain and execute arbitrary code inside the GPU process.

What are you reading this weekend?

Double-speed Flash?

Bizarre computer behavior number 1,413:

After some recent update in the last week or so (I'm not sure which update, specifically), my Flash player is playing all videos at double-speed!

And I can't find any Flash setting or control or widget that lets me control the playback speed.

Does anybody know why Flash would decide to playback videos at double speed, or how I can control Flash's playback speed?

Where does Flash keep its configuration and customization information?

Anand-Gelfand game 11: draw

Gelfand had White, Anand's 7th move varied from game 9, and the result was a 24-move draw. Shipov:

Having a wide choice isn't always a good thing, my friends. You're beset by constant doubts and a lot of time goes nowhere.

Sunday is a rest day; Monday is the final 12th game at classic time control. If Monday's game is not decisive, I believe the match switches into a different format, with games at a much sharper time control.

Friday, May 25, 2012

Another crowded weekend on Mount Everest?

The forecast is good, the teams are climbing, and the MountainTrip team have summited.

Queen Anne Tower Reconstruction in Alameda California

Here's a nice video made by my friend and co-worker Alan about the elegant renovation/restoration work he did on his 120-year-old house's tower.

2010-2011 Reconstruction of the tower on an 1895 Queen Anne mansion. This project won a 2012 Architectural Preservation Award from the Alameda Architectural Preservation Society.

The result is very successful, congratulations Alan!

Stunning pictures of the American west

Even though they're actually only 140 years old, you really should have a look at these gorgeous pictures: In Focus - The American West, 150 Years Ago - The Atlantic.

I particularly like the picture of Virginia City, Nevada, from back in the days when some guy named Mark Twain was living there, writing Roughing It. In many respects, Virginia City still looks similar; if you've never been, it's definitely worth a visit (though don't go now; go March/April or October/November, otherwise the weather is too severe).

Is Elon Musk Tony Stark?

I'm not much of a billionaire-celebrity type, but you gotta admit that what SpaceX is doing is pretty amazing.

Thursday, May 24, 2012

Anand-Gelfand game 10: draw

Dana Mackenzie has his translation of GM Sergei Shipov's commentary on his website

From the opening to the endgame at a full gallop! That’s the way it often is in contemporary theory. How many times have we seen the Berlin endgame alone… I hope that this endgame will be a little bit more interesting. At any rate, it will be a painting on a blank canvas. Without preparatory sketches, without theory and templates.

Wednesday, May 23, 2012

Anand-Gelfand game 9: draw

Dana Mackenzie has posted his translation of GM Sergei Shipov's beautiful and lyric commentary on his blog.

This is the psychology of our Soviet people. When things are at their worst and the tension is so thick you can cut it with a knife, then we are strong and as tough as nails. But when everything is all unbelievably good, when we are showered with money and have plenty to eat and everything else we want, then we start to worry. It’s like the classic song of Boris Grebenshchikov: “Where everything is good, they’ll stab me in the heart. Where everything is crappy, that’s where I feel at home…”

Software patents

I'm (temporarily) distracted from chess news by all the big software patent news coming out this week:

The verdict is apparently in on the Google-v-Oracle patent lawsuit over Java: Google was cleared of patent infringement. Groklaw is covering the developments here.

And the Supreme Court has made a most interesting decision in the Ultramercial-v-Hulu case; namely, asking the Federal Circuit to reconsider the ruling in the light of Mayo-v-Prometheus. The EFF has a writeup here.

I'm sure there will be a lot more to read about these matters in upcoming weeks as everyone sifts through and digests the rulings.

Meanwhile, back to chess...

Tuesday, May 22, 2012

Chess even when there's no chess

Did I mention that today is a rest day in the suddenly-quite-exciting world championship?

Well, that's no reason to sit around with a long face; there's plenty of other interesting chess news that was on my mind:

A nice article from Dana Mackenzie about why translating is still more art than science: Of Knights and Knaves on F5 ... or Why Translations Will Never Be Automated

If you ever want to think deeply about a language (even your own language, or perhaps I should say especially your own language), you should try translating something into it or out of it.

Vinay Bhat reflects on the match, at its halfway point (before the most recent two decisive games): At the (Classical) Halfway Point

But just like it’s easier for me to suggest that my opponent sacrifice his pieces, it’s easy to tell two players playing for the World Championship that I want them to take more risks and possibly lose.

At my Alma Mater, a casual speed chess culture has developed: In back of Hutch, a nightly game of kings

“You can actually learn a person’s character by seeing him playing,” he said. “Stingy people don’t like to sacrifice a piece. But sometimes you need to do so in order to win. You can also tell whether a person is aggressive or materialistic in the same way.”

And a nice follow-up piece highlighting Nakamura's final game in the U.S. Championship: Nakamura wins U.S. Championship

The critics (including me) would have said "What was he thinking!" if Naka had lost. But no, it worked like a charm - Naka crushed Seirawan, while Kamsky drew. This gave Naka the championship by a full point, undefeated, 8.5/11 (6 wins, 5 draws). This is comparable to Fischer's typical score in U.S. Championships. Congratulations to Hikaru Nakamura on his gutsy play and incredible performance!

You really should play through that Nakamura-Seirawan game -- it's magnificent! Seven of Black's first 11 moves are pawn moves, and by move 15, when White has 6 developed pieces to Black's one, it's hard to believe you're watching a current and a former U.S. Champion playing each other; it looks like one of those games you might encounter in Hutchinson Commons in the Reynolds Club.

Flummoxed juries

It's a rest day in the world of chess, so we can return to the world of technology.

Today, we take a trip over to the still-underway clash-of-the-titans intellectual property lawsuit between Oracle and Google over the use of the Java programming language on Android smart phones.

Caleb Garling of Wired Magazine has been following the trial closely, and filed this report on the Wired website today: Jury Flummoxed Over Google-Oracle Patent Fight.

In the article, Garling describes how the court is handling questions of technology:

The jury has been deliberating over the claims for a week now, and on Tuesday, it had two more questions for the court, and both were related to the nuances of “symbolic references” and how they apply to data retrieval.

First, the jury asked whether the ’104 patent covers a symbolic resolution anywhere in the data fields during compilation, and Judge Alsup said that it does not specify a place. Later in the morning, the jury asked whether the use of numeric memory locations in Android compilation precluded the existence of symbolic references.

Judge Alsup turned to lawyers from both sides and asked them to agree on the wording of the answer. Oracle said the answer should be “no” and Google said it should be “yes.” Typically, the sides will negotiate on questions like this before reaching the answer, but recently, as the jury has taken longer and longer to reach a verdict, Alsup has grown increasingly frustrated over how much time this takes.

“Once again, the lawyers are of zero help,” he said. “The sides don’t agree on anything. So I have to figure it out on my own.”

He then called in the jury and gave his best explanation. Judge Alsup revealed last week that while he didn’t know the Java programming language before the case started in August of 2010, he had taken it upon himself to learn the basics. And he indicated that he had coded in other languages prior to the case.

“If you find a numeric reference, that’s a numeric reference. It can’t be both,” he said, clarifying that when examining the references, the jury must examine them in on case by case basis. “For any line item, it’s either numeric or symbolic…But if it’s numeric way [in one place], [another place] might be symbolic.”

In answering both questions for the jury, Judge Alsup seemed frustrated. “It’s not easy for me. And it’s not easy for you. Now go back in the jury room and dig back in,” he said as the jurors resumed their deliberations.

The mind reels at the prospect of this technique for resolving disputes.

This week on Mt Everest

There's lots of coverage of the events of last weekend on Mount Everest.

Outside Magazine's reporting has been superb; let me recommend this article as a starting point: FOUR CONFIRMED DEAD IN TWO DAYS ON EVEREST: Disaster strikes world's highest peak with only human error to blame.

Even now, two days after the chaotic events, the details are foggy. That's because of inherently poor communications and the fact that many climbers are so exhausted and woozy from their efforts at altitude that they have a hard time even remembering what happened during their own climbs, let alone those of their teammates and strangers. With radio communications further hampered by geology and an endless stream of information that’s difficult to verify, it would be easier to report on a moon landing.

My close friend Trudi, who was attempting this climb, but was not on one of the teams that climbed last weekend, has chosen to end her attempt and is on her way home. I'm sure she is sad to miss the summit, but we'll all be ecstatic to have her safely home again.

Monday, May 21, 2012

Comparing the Coursera and Udacity Cryptography classes

Over the spring, I worked my way through two online cryptography classes:

Taking both classes was not a waste; the two presentations nicely complemented and reinforced each other, and in the end I felt that I learned a lot from each.

But there were both differences and similarities between the courses, which I found interesting.

Both formats are broadly the same:

  • Both classes are, roughly, upper-division undergraduate classes in modern cryptography, covering roughly the same material
  • Both classes are presented as video lectures, in the style of allowing you to view the virtual whiteboard that the lecturer is drawing on, while listening to the lecture.
  • Both classes augment the video lectures with quizzes, homework assignments, additional reading, and forums for students to gather and discuss the material

Some of the differences in the classes were mechanical:

  • The videos for the Boneh class were generally 12-20 minutes in length, each, while the videos for the Evans class were generally 1-3 minutes in length, each
  • The slides and quizzes for the Boneh class were typeset, and were supplemented by a certain amount of free-hand annotation, while the slides and quizzes for the Evans class were presented entirely in free-hand
  • The classes present the material in a different order
  • Although the materials used largely similar notation, there were notational differences (e.g., Boneh uses the terminology "public key/secret key" and the letters "p" and "s", where Evans generally uses the terminology "public key/private key" and the letters "u" and "r")

The most important differences, though, are a bit harder to describe.

I felt that the Boneh class was more theoretical, more rigorous, and more oriented around the mathematical aspects of cryptography. Boneh emphasized a fair amount of probability theory and number theory during the class, spent more time proving theorems during the videos, and used more structured techniques such as the probabilistic "adversary games" that I wrote about several weeks ago.

Meanwhile, I felt that the Evans class was more practical, more approachable, more intuitive, and more oriented around the applied aspects of cryptography. Evans uses many examples from real life, concentrates more on results and less on derivations, and presents the material in context of its use. Evans also spends a great deal more time discussing the history of cryptography research, with interesting illustrations of important milestones and events, pictures of the people and objects that have been part of the history of cryptography, and so forth.

In general, I was extremely impressed by both courses. The material was clear, accurate, relevant, well-presented, and thoroughly explained. A motivated student who approaches either of these courses with energy and time and commitment will reap significant rewards. After just a few months of serious study, I felt like I had substantially improved my knowledge of these areas, and felt both interested in and prepared to take my studies further.

I hope that this is the beginning of a widespread availability of educational materials of the finest quality on the Internet, and I hope that these courses both find the audience that they deserve.

Update: Professor John Regehr provides a look at the Udacity environment from the other side of the desk, describing his work preparing a class on Software Testing for Udacity

it became clear that designing good programming quizzes is one of the keys to turning lecture material into actual learning. Tight integration between listening and hacking is one of the reasons that online learning will — in some cases — end up being superior to sitting in class. So this became my main challenge: creating a sequence of programming quizzes that basically leads students through the material.

Anand-Gelfand game 8: Anand wins to level the score

Well, I guess that yesterday's result has awoken the world champion, as today Anand won in an almost-inconceivable 17 moves!

It seems that the players moved into un-prepared opening territory quite quickly today, and Gelfand moved aggressively to win Anand's Rook at the price of a Knight, but failed to foresee that his queen was then trapped.

GM Sergei Shipov has this to say about Gelfand's apparent novelty on move 7:

I must admit I wanted to write about this alternative for Black, but I thought it was insufficiently solid - not something for a World Championship match. Black's provoking a white pawn avalanche i.e. the obvious move now is the g2-g4 advance, after which it's hard to come up with anything other than retreating to f6. And what do you get? Black will have lost one and a half tempos and given White the basis for an attack. An extremely risky strategy from the Challenger! As it's not hard to guess, this is a novelty.

The second half of the world championship is certainly much different from the first half; what will happen next?!

A fine spring Saturday at Tomales Bay

Setting off at 9:00 AM on a perfect Saturday morning for kayaking on Tomales Bay with my 8-year-old granddaughter:

We kayaked up to Shell Beach in Tomales Bay State park, where we took a break to gather shells and make sand castles:

Hannah says kayaking is "OK"; perhaps we will do it again one day soon!

Sunday, May 20, 2012

Librarianship questions I wonder about...

... include: when a patron wanders into your library and says "I've read everything that Donald Westlake wrote, and I want more. Is there anything else?", do you give them some of those Richard Stark novels?

(Same question when they ask if that guy Evan Hunter ever wrote anything worth reading...)

Anand-Gelfand game 7: a win for Gelfand!

Wow! The game started just like all the previous games in which Gelfand had White, but explodes at the end with a marvelous mating net of the Rook and two Knights; at the end, Anand queens his pawn but is mated.

GM Sergei Shipov's comments on the game are delightful and sharp:

In such a lamentable position there is no good advice for Black.

...

It's painful to look at the bishop on c8 ... As, indeed, it is painful to look at the world champion himself on the video feed. Black holds on, which is to say that he prolongs the resistance by playing forced moves. Meanwhile White attacks with moves that can easily be made by the hand alone, without the brain's assistance.

A wonderful game, beautiful and violent; surely we will now see much more fighting chess this week!

At any rate, the score now stands: Gelfand: 4, Anand: 3.

And, meanwhile, over the weekend Nakamura beat Kamsky for the first time ever (and with Black, no less), to take the lead of the U.S. Championship, and followed that up the next day with a quick victory over Seirawan to take the title. Congratulations to 2012 U.S. Champion Hikaru Nakamura!

Friday, May 18, 2012

Anand-Gelfand game 6: draw

The score is now 3-3, halfway through the scheduled 12 games.

As GM Sergei Shipov notes, the whole world is trying to figure out when a decisive game might occur, and how the evolution of the match favors the one player or the other:

In the commentators' studio now on the official website Kasparov is talking about the current condition of the players in the match. Unfortunately [or not so unfortunately for most readers here!], only in English... The ex-World Champion said that Gelfand's title chances increase with each new draw... Garry is now comparing the current match to the previous friendly match in Zurich between Kramnik and Aronian. His assessment wasn't hard to predict... At the same time, he noted the difference in the status of the matches, explaining that without the burden of responsibility you can of course play more freely and riskily - as Vladimir and Levon did. While Vishy and Boris can't allow themselves such liberties.

The World Championship has a day off tomorrow, so take a break from studying those Slav Defense Meran variations and wander over to the U.S. Championship to see the thrilling finish between GMs Kamsky and Nakamura, two of the top 10 players in the world going neck-and-neck for the championship.

You can find Kamsky's superb win over Seirawan here.

Thursday, May 17, 2012

Anand-Gelfand game 5: draw

Dana Mackenzie has the translation of GM Sergei Shipov's analysis here.

Do you hear the murmurs of discontent in our ranks? The public is worried, it demands decisive games in the world championship match. Or at least a bitter struggle to the last drop of blood, to bare kings!

Fast money has smelt blood

Are you, like me, struggling to figure out what the whole "London Whale" event has been about, racing to try to catch up even as things seem to worsen daily?

I'm certainly no expert in this area, just an interested observer, but I thought I'd share what little I'd learned.

First, start with this fine introduction from Heidi Moore at USC's Marketplace site. Moore explains the overall story at a high level, without getting bogged down in detail:

JP Morgan also owns a lot of "high-yield," or junk bonds. Those are low-rated bonds from companies that are close to default. If JP Morgan owns a lot of high-yield bonds, the best way to balance that risk is to bet on investment-grade corporate bonds.

Iksil wanted to find a way to bet that those investment-grade bonds were totally secure, that they would never default.

When you're ready for a more detailed and intricate description, find your way to Lisa Pollack's thorough explanation at the Financial Times Alphaville website: . As Pollack explains the situation, it involves various large market participants taking different positions on the trading:

  • The mystery trader, perhaps Mr Iksil, has gone long in a big way on the Markit CDX.NA.IG.9, i.e. sold large amounts of protection against the 121 credits contained therein.
  • Hedge funds are aware of the big position.
  • The hedge funds are using skew trades (we’ll explain shortly) to express this view.

Just above those paragraphs though:

J.P. Morgan said the CIO unit is “focused on managing the long-term structural assets and liabilities of the firm and is not focused on short-term profits.”

The bank added, “Our CIO activities hedge structural risks and invest to bring the company’s asset and liabilities into better alignment.”

So as far as JP Morgan is concerned, these are long-term hedges. So, reputational risk aside, are these really positions to bet against? After all, if the $100bn position that was referred to in the Bloomberg piece is gross notional, that sounds about right for the size of JP Morgan’s exposure anyway. As of March 30, 2012, there was $884bn gross in total on the index (untranched), and JP Morgan is a huge player in this space. So, whatever… big for sure, but not like crazy mental big.

Skew trading is where one buys or sells protection on the index, while simultaneously selling or buying an equivalent amount of exposure to the underlying credits. All 121 of them in the case of the CDX.NA.IG.9.

At the moment, what the hedge funds that spoke to the journalists are saying is that Iksil has supposedly sold so much protection on the index that its spread is a lot lower than it should be relative to the single-names. So they are betting that the skew will have to go back towards zero (it’s negative now).

Is it sounding like a plot line from The Sopranos yet?

For still more detail, you might find Jon Macaskill's reporting at the Euromoney site interesting, for example: Inside JPMorgan's $2 billion loss-making CIO division. Macaskill says:

The size of the value at risk of the CIO portfolio is startling and has exceeded the VAR in JPMorgan's investment bank in each of the past two quarters, which indicates that Dimon must have approved the risk levels being run by the unit. The CIO had average VAR of $69 million in the fourth quarter of 2011 and $67 million in the first quarter of this year. JPMorgan's investment bank - nowadays the most successful in the world in terms of revenue generation - managed to get by with VAR of $57 million in the fourth quarter of 2011 and $63 million in the first quarter of 2012.

VAR, I believe, is "Value At Risk". The CIO is the "Chief Investment Office", not the (more common in my industry) Chief Information Officer. The JPMorgan CIO seems to have had not enough information, and too much at risk.

Lastly, you'll certainly want to keep up-to-date by monitoring the incredibly prolific Felix Salmon of Reuters, who has written a series of fine articles on the situation, including:

Salmon summarizes the event as follows:
There are two things you can do when something starts to go wrong in the markets. You can unwind your position at a loss. Or you can try to fix it. Iksil, and Drew, chose the latter.

I'm trying to learn about these things, to be an informed citizen, but it's hard. The "financial engineers" at these large banks have cloaked themselves in mystique and intrigue, and, except for those times where they declare themselves "too big to fail" and demand a few hundred billion from the world's taxpayers, they'd prefer to be let alone to make (or lose) their billions in secrecy.

Still, if you have any good references to sources that explain and illuminate this strange incident, by all means send them along!

Tuesday, May 15, 2012

Unwieldy companies overrun with vice presidents

The Yahoo epitaphs are starting to come fast and furious nowadays, unsurprisingly.

Don't miss this epic article over at Gizmodo: How Yahoo Killed Flickr and Lost the Internet. Some of it is the typical sort of combination of sour-grapes griping and revisionist history that these articles tend to bring, but this one is well-written and has some genuine insights and some strong lessons learned. For example:

Yahoo needed to leverage this thing that it had just bought. Yahoo wanted to make sure that every one of its registered users could instantly use Flickr without having to register for it separately. It wanted Flickr to work seamlessly with Yahoo Mail. It wanted its services to sing together in harmony, rather than in cacophonous isolation. The first step in that is to create a unified login. That's great for Yahoo, but it didn't do anything for Flickr, and it certainly didn't do anything for Flickr's (extremely vocal) users.

Yahoo's RegID solution turned out to be a nightmare for the existing community. You could no longer use your existing Flickr login to get to your photos, you had to use a Yahoo one. If you did not already have a Yahoo account, you had to create one. And you did not even log in on Flickr's home page, upon arriving, you were immediately kicked over to a Yahoo login screen.

This sort of merger-and-acquisition pain is all too real; I've been through plenty of it myself. It's a heartbreak, as the quote above shows, for it's always full of well-meaning individuals trying to do the right thing, but even though they have the best ideas and the best intentions, what results is catastrophe.

One of the best aspects of the article is the way it analyzes how the various decisions that were made along the way impacted that most delicate and crucial concept: the community:

While other apps draw users into their Web services (think Foursquare, Twitter, Facebook, and notably Instagram) the Flickr app that Yahoo Mobile rolled out had no mechanism for that. It was not a recruitment tool. It was just for existing users.

Of course, it's extremely hard to understand, predict, anticipate, and adapt to the changing tastes of the online public:

The story of Flickr is not that dissimilar to the story of Google's buyout of Dodgeball, or Aol's purchase of Brizzly. Beloved Internet services with dedicated communities, dashed upon the rocks of unwieldy companies overrun with vice presidents.

As a result, Flickr today is a very different site than it was five years ago. It's an Internet backwater. It's not socially appealing.

Of course, just to be clear, I'm perhaps the most dinosaur of them all, for I still store my photos (those few of them that I actually take) on Picasa (gasp!).

I mean, how 10-years-ago is that?!

Anand-Gelfand Game 4: Draw

This game seemed quite similar to game 2, in my naive eyes.

Sometime later today, Dana MacKenzie should have the translation of GM Sergei Shipov's coverage on his blog.

While you wait, you can content yourself with all the various links in GM Ian Rogers's nice article: A Couch Potato’s Guide to the 2012 World Championship.

So, once you have followed the games online, eaten the potato pancakes, viewed the press conferences, had lunch (the leftover potato pancakes), then watched the videos and read the pundits' opinions of each game, please go outside and do some exercise or you will end May with a vitamin D deficiency and seriously overweight!

Monday, May 14, 2012

Somewhere in the swamp between Art and Sport ...

... resides LP record Parkour: Vinyl Throw: MusicBunk

Anand-Gelfand Game 3: Draw

But far from a boring draw!

GM Sergei Shipov concludes his comments about game 3 as follows:

The Champion reacted extremely successfully to the Challenger's prepared idea in the opening, seized the initiative, applied pressure, was close to victory, but didn't have enough stamina or time. An irony of fate - once the world's fastest player fell victim to time trouble. Times have changed and Anand's no longer so young.

Saturday, May 12, 2012

Anand-Gelfand game 2: draw

Sergei Shipov feels that Anand equalized with 14. ... Nf6.

Meanwhile, Vinay Bhat, one of the strongest local players in my area, provides a nice preview article with his thoughts, and several other articles to read.

Anand’s repertoire is much broader, making preparation much more difficult. Gelfand of course will be well prepared, but he will have had to do work on 1.d4, 1.c4, and 1.e4 (all of which Anand has played against him). With colors reversed, Anand doesn’t have quite the same dilemma – the odds of Gelfand playing 1.e4 in this match are precisely 0.

Friday, May 11, 2012

Stuff I'm reading on a Friday afternoon

A couple interesting things to keep you at your computer on a warm spring day!

  • Kathie Nichols and Van Jacobson's recent article in ACM Queue has spurred a lot of interesting discussion. Here's some of what I've seen so far:
    • Jim Gettys is hailing the work as Fundamental Progress Solving Bufferbloat, and notes that
      A preliminary Linux implementation of CoDel written by Eric Dumazet and Dave Täht is now being tested on Ethernet over a wide range of speeds up to 10gigE, and is showing very promising results similar to the simulation results in Kathie and Van’s article.
    • Not everybody is so happy about the new queue management proposal. Bram Cohen, in particular, lit into it with a strongly-worded article, TCP Sucks
      For the end user to complain about how big the buffer is would be like them complaining to their credit card company for offering too high of a limit. ‘You should have known I’d spend too much!’ The solution is for the end user to intervene, and tell all their applications to not be such pigs, and use uTP instead of TCP.
    • Avery Pennarun added his own thoughts in an interesting article: TCP doesn't suck, and all the proposed bufferbloat fixes are identical.
      Oddly enough, fixing TCP to work around bufferbloat is pretty easy. The solution is "latency-based TCP congestion control," the most famous implementation of which is TCP Vegas. Sadly, when you run it or one of its even better successors, you soon find out that old-style TCP always wins, just like it always wins over uTP, and for exactly the same reason. That means, essentially, that if anyone on the Internet is sharing bandwidth with you (they are), and they're running traditional-style TCP (virtually everyone is), then TCP Vegas and its friends make you a sucker with low speeds. Nobody wants to be a sucker.
    • And Matthew Sackman from the RabbitMQ team has yet another perspective: Some queuing theory: throughput, latency and bandwidth.
      it's possible to see reasons why CoDel isn't as appropriate for dealing with AMQP messages as it is for plain IP. It's also worth remembering that requeuing messages via nacks is a fairly expensive operation, so it's a good idea to set the parameters of CoDel to ensure in normal operation very few if any messages are being nacked.
  • Separately, I've also been looking through a series of recent articles by the team at BlueBox.net:
    This document was written with the goal of giving you a place to start to understand the concepts and concerns involved, as well as to give some practical advice as to “where to start” if you are trying to turn an application which exists happily in a single datacenter into an application which exists happily spread across two or more data centers.

See, there you go! More things to read, to get smarter!

Anand-Gelfand game 1: draw

If you are following the games, don't mess around: go directly to Grandmaster Sergei Shipov's commentary at the always-wonderful Chess in Translation site.

Thursday, May 10, 2012

Hulk! SMASH!

What, you haven't seen The Avengers yet?

What are you waiting for? It's great!

Mark Ruffalo is superb. The Iron Man movies always made me squirm because Robert Downey Jr. was just a bit too over-the-top for me; Ruffalo is the perfect complement, and watching the two gently toss things back and forth in the laboratory was delightful.

It's not just a game ...

... it's a role model for people trying to design learnable software:

This is about how these two universes should collide and that means what I’m really talking about is gamification. There’s a reason I didn’t mention this until paragraph 17 because there are a lot of folks who think gamification means pulling the worst aspects out of games and shoving them into an application. It’s not. Don’t think of gamification as anything other than clever strategies to motivate someone to learn so they can have fun being productive.

I've always thought that one of the great insights for people trying to learn how to build better software was to realize that great computer games didn't give you a manual. If you start out by giving them a manual, you've failed. As Lopp points out, this applies to all software, not just to games:

That’s how I want to learn. Don’t give me a book; I don’t want a lecture, and I don’t want a list of topics to memorize. Give me ample reason to memorize them and a sandbox where I can safely play. Test me when I least expect it, shock me with the unknown, but make sure you’ve given me enough understanding and practice with my tools that I have a high chance of handling the unexpected.

"Shock me with the unknown" -- what a great slogan!

I remember the first time I was faced with the Chrome browser, and my first thought was: "that's it?" There didn't seem to be enough buttons, and enough places to type, and enough controls to operate the thing. But there was just enough there to get me interested enough to type something, and click on something, and pretty soon I was off and running.

That's gamification.

Wednesday, May 9, 2012

Chess this week

Of course, all the attention is on the Anand-Gelfand match, which begins tomorrow in Moscow.

But don't forget that this week is the 2012 U.S. Chess Championship, being held in Saint Louis at the Chess Club and Scholastic Center of Saint Louis.

The championship is an invitational tournament, with 12 of the top players in the U.S. invited, including the two strongest players in the country, Gata Kamsky and Hikaru Nakamura. There's a women's tournament, too, and it also has a very strong field.

Round 1 was a doozy, with multiple decisive games. Follow the games live here!

Monday, May 7, 2012

The Politics of Competitive Board Gaming Amongst Friends

Here's a fun 10-minute film: The Politics of Competitive Board Gaming Amongst Friends.

If you've ever been part of a gaming group, the whole thing will strike you as "true, too true."

And if you're like me, you'll probably recognize little bits of yourself (and your friends) in each of the various characters.

Sunday, May 6, 2012

Learn he did, not always happily

Let me call to your attention Lawrence Downes's column in today's Times: Back to ‘Graceland’ After 25 Years.

This year is the 25th anniversary of many things, but one of those things is Paul Simon's astonishingly perfect album Graceland.

As part of the anniversary, a new documentary is out: Under African Skies. In the movie, Simon returns to South Africa to reflect and discuss on the events of that year, and his part in them.

I was just coming of age during those times, old enough to support the boycott, and understand that what was happening in South Africa was bad, and needed to change, and needed to change now.

It's always hard to discuss the role of artists in situations like this: was Simon wrong to do what he did? His work brought attention to the situation, after all. Downes notes:

“Graceland” was his biggest hit, a global phenomenon, but success made it a target. Anti-apartheid protesters picketed the world tour. African-American college students accused him of doing what white musicians always do when they find black music that’s irresistibly good. Boycott leaders deplored the thought of an outsider sidestepping their global campaign without permission or apology, and bringing a group of South African artists along with him.

Perhaps most importantly, Graceland still, 25 years later, stands as a work of art, and it has stood that test well. Downes concedes:

The regime is dead; apartheid’s defenders are gone and forgotten. Reagan is remembered for other things. Righteous anthems like Peter Gabriel’s “Biko” are trapped in their time.

But “Graceland” is not. It’s as alive and surprising now as it was then, when South African artists shared their genius with Mr. Simon.

I am of course an enormous fan of Paul Simon; he is one of the greatest talents of my time. I will look forward to watching "Under African Skies" when I get the chance.

Saturday, May 5, 2012

Neonicotinoids and Colony Collapse Disorder

For many years I have been interested in and following the investigations into what is causing Colony Collapse Disorder.

It's been well more than 5 years of intensive study now, trying to figure out what affects the bees, and why.

I recently read an article in The Economist: Subtle Poison, which reported on several recent developments in the investigation:

Many researchers believe the label “colony collapse disorder” covers a multitude of problems; that would account for the long list of possible causes. But neonicotinoids have the explanatory virtue of being a fairly recent development and also one which, as these two pieces of work suggest, could be a common factor in weakening a colony without actually pushing it over the edge. The killer blow would then be administered by something else: a mite infestation, perhaps, or a fungal infection, or whatever else happened to turn up that a healthy hive would have shrugged off. A paper published earlier this year in Naturwissenschaften, for example, showed that even small doses of neonicotinoids weakened bees’ resistance to Nosema, a common fungal parasite.

The recent developments were published in Science Magazine, and include:

I love the part in the second paper about "labeled with a radio-frequency identification tag"; as The Economist article observes, this involved an interesting field experiment:

To that end, he and his colleagues glued tiny radio transmitters to the thoraxes of worker bees. These triggered a detector on the hive whenever a worker bearing one returned from a foraging trip. Some hives were given realistic doses of thiamtethoxam, a variety of neonicotinoid, while others were left alone. Dr Henry found that around twice as many treated bees as untreated ones failed to return to the hive.

I can tell you for sure that one task I wouldn't be very good at is gluing tiny radio transmitters to the thoraxes of worker bees!

What's the bottom line? Well, as the American Chemical Society points out: Pesticides Harm Hive Behavior. That's pretty simple, and now it seems that it's quite well-established.

And The Economist notes that countries are starting to take action:

A few countries, including France, Germany and Slovenia, have already restricted the use of neonicotinoids because of worries about their effects on bees.

Hopefully more countries will follow suit, and we can halt the poisoning of the bees soon.

Friday, May 4, 2012

Stuff I'm reading on a Friday afternoon

When compile times bog down a bit, and there's the occasional minute for reading, I find myself reading some stuff on the Internet. Here's a sample from today:

  • Probabilistic Data Structures for Web Analytics and Data Mining
    In this article, I provide an overview of probabilistic data structures that allow one to estimate these and many other metrics and trade precision of the estimations for the memory consumption. These data structures can be used both as temporary data accumulators in query processing procedures and, perhaps more important, as a compact – sometimes astonishingly compact – replacement of raw data in stream-based computing.
  • Replicated/Fault-tolerant atomic storage
    There is this elegant algorithm for replicated/fault-tolerant atomic storage that I think every distributed systems researcher/developer should know about. It is simple and powerful. And, it is fun; I promise your brain will feel better about itself after you learn this majority replication algorithm.
  • Common statistical fallacies
    I've been reading papers on how people learn statistics (and thoughts on teaching the subject) and came across the frequently-cited work of mathematical psychologists Amos Tversky and Daniel Kahneman. In 1972, they studied statistical misconceptions. It doesn't seem much has changed.
  • Bell’s-inequality-denialist Joy Christian offers me $200K if scalable quantum computers are built
    Like Gödel’s and Cantor’s Theorems, Bell’s Theorem has long been a lightning rod for incomprehension and even anger; I saw another “disproof” at a conference in 2003, and will doubtless see more in the future.
  • Notes on graph data management
    Interest in graph data models keeps increasing. But it’s tough to discuss them with any generality, because “graph data model” encompasses so many different things. Indeed, just as all data structures can be mapped to relational ones, it is also the case that all data structures can be mapped to graphs.
  • More fun with DateTime
    There's one piece of inherent date/time complexity you'll need to understand for this post to make sense: sometimes, a local date/time occurs twice. For the purposes of this post, I'm going to assume you're in the UK time zone. On October 28th 2012, at 2am local time (1am UTC), UK clocks will go back to 1am local time. So 1:20am local time occurs twice - once at 12:20am UTC (in daylight saving time, BST), and once at 1:20am UTC (in standard time, GMT).
  • Common Lisp: The Untold Story
    This paper summarizes a talk given at “Lisp50@OOPSLA,” the 50th Anniversary of Lisp workshop, Monday, October 20, 2008, an event co-located with the OOPSLA’08 in Nashville, TN, in which I offered my personal, subjective account of how I came to be involved with Common Lisp and the Common Lisp standard, and of what I learned from the process.

    The account highlights the role of luck in the way various details of history played out, emphasizing the importance of seizing and making the best of the chance opportunities that life presents. The account further underscores the importance of understanding the role of controlling influences such as funding and intellectual property in shaping processes and outcomes. As noted by Louis Pasteur, “chance favors the prepared mind.”

Thursday, May 3, 2012

Serious online education seems to finally be taking off

Between Udacity, Coursera, and now edX, serious online education seems to finally have some momentum.

I've been taking a Cryptography class at Coursera and it's been superb. I've also dabbled in a Coursera Algorithms class and it is also very good.

I haven't taken any of the classes at Udacity or edX, so don't have much to comment on there.

What's not clear to me is why there are suddenly multiple such initiatives. Why did MIT and Harvard feel they had to start their own initiative, rather than just joining one of the existing ones?

At any rate, it's hard to look such fine gift horses in the mouth, so I shan't. Thank you, modern institutions of higher education, for making these resources available to the world!

As if I didn't already have enough claims on my time...

Skyrim Goes MMO: Bethesda Announces Elder Scrolls Online

Of course, MMO has never been my thing, so I'll probably end up passing (he says to himself...).

Wednesday, May 2, 2012

I am a sucker for lists

I love lists.

I'm not sure why, but I love almost any sort of list, whether it be a checklist, a catalogue, a best-of list, or whatever.

So, without further ado, 3 interesting lists (at least, they were interesting to me):

  • I can pretty much guarantee that you will find something fascinating that you had overlooked before in Conor Friedersdorf's 101 Spectacular Nonfiction Stories
    Each year, I keep a running list of the most exceptional nonfiction that I encounter while publishing my twice-weekly newsletter The Best of Journalism. Along with my curating work for Byliner, this hoovering of great stories affords me the opportunity to read as many impressive narratives as any single person possibly can. The annual result is my Best of Journalism List, now in its fourth year.
  • A rather practical list showed up recently in the OCSC sailing blog: “Rockin” in the free Bay
    we can address the interesting query from a reader about where the ‘hazard to navigation’ rocks are in San Francisco Bay, and which ones sailors should be particularly concerned about. Oddly, this is the first time we’ve ever gotten this question and it took a bit of head scratching to come up with a list, since not all of the ‘hittable’ rocks are noted on charts, and not all notable rocks are hittable. Anyway, here’s what we came up with.
  • And, lastly, the always-fascinating Roger Ebert revisits (for the fifth time) his list of The greatest films of all time
    I decided not to do that--trash the 2002 list and start again. It was too much like a stunt. Lists are ridiculous, but if you're going to vote, you have to play the game.
  • Update: And here's a rather sad list:
    David Griggs died in a car wreck while driving drunk. Rodney Culver died in the crash of ValuJet Flight 592. Doug Miller was struck by lightning. Curtis Whitley died of a drug overdose. Chris Mims died of an enlarged heart, Shawn Lee of cardiac arrest, and Lew Bush of a heart attack. And on Wednesday, 43-year-old Junior Seau became the eighth member of the 1994 San Diego Chargers to pass away before his 45th birthday.

Seen any great lists lately? Let me know!

Tuesday, May 1, 2012

Microsoft C++ OMFG

Hey, all you people who design and implement programming languages, listen up for a second!

If you find yourself building a feature in your programming language that necessitates the production of materials like this and this to explain it to the developers using your language, and if you find those developers experiencing problems like this and reporting bugs like this, then

STOP

You have gone way beyond the bounds of sane behavior and need to find somebody to bring you back somewhere into the same solar system as reality.

Just saying.

NSDI 12

Usenix continues its streak of running the most interesting and relevant conferences.

This time it's NSDI 2012, the 9th USENIX Symposium on Networked Systems Design and Implementation, that was held last week.

In additional to being topical and relevant, Usenix also leads the way in providing open and widely available access to their work. The conference is just over, and they have made the conference materials available to all. Bravo! (I'm pleased to see that, finally, some of the academic community appear to be catching the wave, as well.)

Among the papers that caught my eye as I studied the technical sessions were:

There are other papers that look intriguing as well; as I say, these were just some of the ones that caught my eye right away.

I'll have more to say about several of the papers later; for the time being, it's time to do some reading :)