Kirk Pearson's Weblog

A 4,000 Megapixal Camera A recent Wired Magazine article describes a film camera, designed by physicist Graham Flint from spy plane and nuclear reactor parts, with a resolution of 4 gigapixels. The camera uses 9x18 inch photographic plates with a resolution of 4,000 pixels per inch. Images from the camera are scanned with a Leica Geosystems scanner, which produces 4 gigapixel digital images. Each image fills one DVD. "The images are printed in strips on a large-format Epson 9600 printer and mounted on panels like wallpaper." One print could measure up to 48x24 feet. Flint designed the camera to "photograph the entire Milky Way in color and very high resolution," but his New Mexico observatory was shut down for health reasons before he could begin that project. He now tours the United States making thousands of photographs of "cities, monuments and national parks." You can see more of his pictures at his Gigapxl Project website. (2005-02-07 15:35:09.0) Permalink Comments [1]

Sony will launch smaller PStwo on October 26

According to a Yahoo! News article, Sony is expected to launch the new PStwo console in the U.S. on October 26. The new console is 30% smaller than the current PlayStation 2, and should retail for the same US$149 price. In other news, Take-Two Interactive has delayed the release of its "Grand Theft Auto: San Andreas" game by one week to coincide with the launch of the PStwo, inciting Grand Theft Auto fans everywhere to riot. Thanks a lot, Sony!

The original Star Wars movies are now in High Definition

But you won't be able to see them that way for a while. A September 6, 2004 USA Today article describes the digital restoration process the original Star Wars movies have undergone to prepare them for their first release on DVD (on September 21, 2004). Original movie prints get scratch, scuff and dirt damage every time they are copied, and the Star Wars prints were in pretty bad shape. A company called Lowry Digital Images has digitized each frame of the original prints and retouched them using "80 employees and 600 networked Power Mac G5 computers with the equivalent of 378 terabytes (378 million megabytes) of hard-disk storage" to remove the signs of wear and tear, and now "the footage seems to shine, as if brand-new." (I am picturing the brand-new shine of C-3PO at the medal ceremony at the end of Star Wars :-))

Is digital restoration a good thing for classic movies, or does it change them for the worse in some fundamental way, like colorizing black & white movies? It sounds like the good will far outweigh the bad. And George Lucas' editorial changes to scenes in the movies could be far more damaging :-) Hopefully they'll be improvements, like his changes to his first film, THX-1138. I guess we'll see in a week.

Oh, about the High Definition thing: now that the movies are digitized, a High Definition master of them exists, and it can be used to make new film prints and High Definition DVDs.

Genesis augers in

The Genesis return capsule, which was to return particles of the Sun to Earth, has crash landed after its two parachutes failed to deploy. Two helicopter pilots who trained for 6 years to catch the capsule in mid-air, didn't get a chance to catch it as it plummeted to Earth at over 100 MPH. The capsule has been located, at least. A September 5 Wired News article quoted project scientists as saying that if the capsule crashed they would "have to spend months sorting through broken jewelry-studded disks holding the tiny solar wind particles." From this CNN image, it looks like the capsule may be mostly intact, though.

The helicopter recovery was supposed to be the most difficult part of the mission. Who would have thought the parachutes would be the point of failure? But give NASA credit for trying difficult and innovative missions. Without trying, there would be no progress.

MD5 crypto flaw weakens Sun's Solaris Fingerprint Database

The MD5 encryption algorithm, widely used for things like bank Automated Teller Machines (ATMs), bank electronic fund transfers, the Apache web server, Sun's Solaris Fingerprint Database, and creating secure digital signatures, has been cracked. On August 17, four Chinese researchers released a paper detailing a way to exploit flaws in the MD5 algorithm. A distributed computing project, called MD5CRK began 6 months ago, with the goal of proving that the MD5 algorithm is vulnerable. With the project's most recent level of participation, it determined that it might need 200 years to find a collision of points which would defeat the algorithm. But the Chinese researchers' method can find a collision within one hour. A ZDNet article has more information about the reported flaws in MD5.

The Netflix paradox

Every time I visit the Netflix DVD rental site, to see whether a movie I have returned has been received, or to change the order of my rental queue, I invariably find another recommended movie (or two or five) that I want to add to my queue. I try to shorten my queue by watching movies more frequently, but the more movies I watch, the more I visit the website and the longer my queue grows. And if I don't visit the site, I get a movie or movies I'm not in the mood to watch and it takes me longer to get around to watching it, so my queue doesn't shrink. So, the more I try to reduce my movie queue, the longer it grows.

Netflix must have figured out early on that targeted movie recommendations is a good way to keep customers coming back. And Amazon.com figured out years ago that targeted book and other media recommendations is a good way to keep customers buying new books faster than they can read their old ones. It's no coincidence that my (and probably your) bookshelf increasingly contains more books that I'm going to read than books I have read.

Is this just a basic flaw of human nature: the more we have, the more we realize we don't have, and the more we want? It is just greed? Is the Internet accellerating our greed by giving us access to more information?

Francis Crick, co-discoverer of DNA structure, died today at 88

According to a Yahoo! article, Francis Crick, who co-discovered the double-helix structure of Deoxyribonucleic Acid (DNA, the building-blocks of almost all life) with James Watson in 1953, died today after a battle with colon cancer. It's hard to believe that DNA was first understood only 51 years ago. Much of modern medicine and molecular biology (and even some computer science) is derived from the work of Crick and Watson.

I first learned about Crick and Watson's work in a great book I read in a college Introduction to Biology class, The Double Helix, by James Watson. First-hand accounts of major scientific discoveries are rare, and well-written, interesting first-hand accounts are even more rare.

If you would like to honor Crick and his work, you can run one of the following non-profit screen-saver distributed computing projects on your workstation or PC to contribute to useful cancer and protein-folding research:

• Parabon Computation - colon cancer research and studies to ease the suffering of chemotherapy recipients
• grid.org - searching for cancer-fighting drugs
• Find-a-Drug - searching for cancer-fighting (and other disease-fighting) drugs
• Distributed Folding - researching how proteins "fold and assemble into living cells"
• Folding@home - researching how proteins self-assemble or fold
• Predictor@home - researching ways to predict protein structures from protein sequences

how the web bypasses language barriers

I want the world to know about distributed computing, and to put its wasted computing cycles to good use. But since I only speak English (yes, I'm another one of those lazy Americans who hasn't learned another language fluently because much of the rest of the world has been gracious enough to learn English), I can only tell others about distributed computing in English.

Fortunately, some of those other world citizens who have learned English have also been gracious enough to translate my site into their languages: Hebrew, Italian, Portuguese, German, Czech, and now Chinese (hello to the Equn.com team!). Thanks to these people I can reach more of the world in spite of my langauge handicap. And thanks to the web, I can do this much more easily and quickly than was ever possible before.

Some day Douglas Adams' babel fish may exist, and none of us will have language barriers, but until then we can use some of the increasing number of tools on the web to help us bypass our language barriers. The babelfish machine translator site allows people to translate small amounts of text or web pages among various languages. The translations are not always of high quality, but they are free, fast, and usually understandable (and the mistranslations are often funny). I use this site frequently to read discussion forum postings in French, German, Swedish, Russian, Korean and other languages: without this site I would miss out on these discussions. Wikipedia, an online, "open-source" encyclopedia, allows people to create encyclopedia articles in many languages and to easily translate them into other languages. The Worldwide Lexicon project will create a "network of multilingual dictionary and translation services for the Internet." This project will allow human volunteers to translate words and phrases quickly and accurately among many languages.

I am amazed at how many ways there are now for people to access all of the information on the web, regardless of the language it is written in.

Google search is broken!

As of 10:15 AM MDT (5:15 PM UTC), if you try to perform a search at Google, you will see:

  503 Server Error

  Google    
  Error
 
  Server Error
  The service you requested is not available at this time.

  Service error -27.

I can't think of a time that Google search has ever been broken when I've tried to use it. I wonder if this is related to all of today's news about Google's upcoming IPO? And speaking of IPOs, what was the last company to IPO between $106 and $136 per share? Google must not be worried about attracting investors.

11:30 AM MDT update: according to a Slashdot article, "the latest MyDoom worm variant has caused a bit of an Internet storm. Google, at this time (12:28 EDT), is returning 503 errors on all queries submitted from certain locations."

Thanks to Chris Calkins for pointing out Google's search problems.