Friday, December 31, 2004

Cool product speeds up your Wi-Fi experience

So you're sitting in the Peanut Barrel, mooching off the Wi-Fi signal from Espresso Royale, a couple of doors down. The signal indicator says "Poor" and those Web pages just aren't painting fast enough.

Now there's a solution:

The accidental telethon -- or the Amazon telethon

Amazon is devoting a huge portion of its home page real estate to an appeal to donate to the American Red Cross for relief for tsunami victims. So far they've raised $7,326,939.19 from 107,683 donors.

Charities worldwide are reporting surprising levels of donation, much of it online. Google's helped by adding a link labeled:

Ways to help with tsunami relief

In effect, this is a worldwide accidental telethon. People see the horrifying images on CNN or in the newspaper, and they want to donate. If they've got handy Internet access, they pop over to a familiar site, such as Google or Amazon, to give.

In some ways, the call to action is as explicit as in the Jerry Lewis telethon for muscular dystrophy: CNN shows scenes of devastation, then flashes on screen the address for the Red Cross, Oxfam, etc. At some times, the Web address of the charity - say, - remains on screen as the camera pans over distressed victims.

Some charities' Web sites have melted down under the load. Amazon is donating more than precious home page real estate; they're providing the back-end of a world-class e-commerce site, easily able to handle the strain as millions consider donating, and thousands follow through.

Amazon is promising 100% of its collections on behalf of the Red Cross will go to the charity. (Let us hope that the ARC spends 100% of it wisely.)

The number of donors via Amazon alone now almost equals the estimated death toll. Let's hope the former increases much faster than the latter. Imagine if the number of donors matched the estimated 5 million displaced survivors....

As inspiring as it is to see the world respond, it's not all good news. We're responding to compelling images. Still, as horrible as things are in coastal Asia right now, where there are no television images, we're not moved to respond. See the movie "Hotel Rwanda" for the images we didn't have in 1994, when 800,000 people were slaughtered. In that case, CNN wasn't bringing us 24 X 7 coverage.

Monday, December 27, 2004

If US tsunami center had warned Asia, could thousands of lives have been saved?

Scientists in the U.S. and elsewhere knew the earthquake had hit, and could've predicted a tsunami in the Indian Ocean. But they had no network to warn nations likely to be affected. Could perfect information have saved lives?

From today's LA Times:

The quake was the largest since a magnitude 9.2 temblor struck Prince William Sound, Alaska, in 1964 and was one of the biggest ever recorded by scientists. It triggered the first tsunami in the Indian Ocean since 1883, civil engineer Costas Synolakis of USC said.

Nations along the Pacific Rim participate in a tsunami detection and warning system, set up since the 1964 earthquake sent a tsunami crashing into the Alaskan coast. But nations on the Indian Ocean have not done so.

In fast, the NOAA / National Weather Service detection service in Hawaii issued this warning:

0315 PM HST 25 DEC 2004




ORIGIN TIME - 0259 PM HST 25 DEC 2004




Take special note of the first line: TSUNAMI BULLETIN NUMBER 001 If this is bulletin 001 for 2004, they made it almost to the very end of the year having issued no warnings for the Pacific. Think about that: this little office on a beach in Hawaii has staff monitoring seismographs all the time. (Today, they issued TSUNAMI BULLETIN NUMBER 003 dated 0536 AM HST 27 DEC 2004 reiterating that despite some small activity in the Pacific, Hawaii is not at risk.)

In the wake of such loss of lives and such devastation, people surely will ask why the Indian Ocean region didn't protect itself better. When people start to point fingers at the governments of affected nations, some folks will say "I told you so." From today's New York Times:
Other scientists have voiced similar concerns. At a meeting in June of the Intergovernmental Oceanographic Commission, a United Nations body, experts concluded that the "Indian Ocean has a significant threat from both local and distant tsunamis" and should have a warning network.

But even given perfect detection, did the region have adequate procedures in place to warn and evacuate people on the coasts? Was there a robust civil defense network in every coastal region. You'd have to educate everyone from locals in poor area shacks, to homeowners in luxury neighborhoods, to folks in bars and restaurants, to tourists in high rise hotels. (Can you imagine a placard on the door of your hotel room that begins "In event of a tsunami warning, please run downstairs and head for the hills...)" And all this for a threat that hasn't hit hard since 1883.

So did any warnings get through? In past disasters, ham radio operators have been instrumental in getting vital messages through. Did anyone hear in advance and evacuate?

Some "what if's" come to mind:

-- Some U.S. seismologists report that they tried to call authorities in countries likely to see the tsunami, but they didn't know who to call. What if they had been able to get through? How many lives might've been saved?

-- What if, instead of trying to call the authorities in those nations, the U.S. tsunami officials had called CNN? Could broadcast media have done a better job of reaching those at risk? (I think the answer is without question yes.)

-- Is there any way the Internet could've helped? The Internet is famous for spreading hoaxes rapidly. What if officials had sent the tsunami warning out to every major blogger in the affected countries and elsewhere, with links back to authoritative sites?

Saturday, December 25, 2004

Photos of Pere Marquette 1225: Inspiration for Polar Express

In November we went up to Owosso, Michigan, where one of the last steam locomotives made in the US resides. It's Pere Marquette 1225, and it inspired a Grand Rapids native to write a book you may have heard of, The Polar Express.

A bunch of photos of 1225, inside and out, and of the museum where it resides, are at:

You have to register one time to visit Imagestation, but it's free.

Here's a little tidbit: many working railroad museums have done "Polar Express" tours for years. With the movie about to come out, Time-Warner did something really stupid: they demanded royalties from the museums. These folks operate as a labor of love, and helped promote the book, and will help sell DVDs of the movie when it comes out.

Warner-Brothers backed off on this stupid, onerous, shoot-yourself-in-the-foot demand, but too late for many museums to call their holiday runs "Polar Express."

So the home of the actual engine that started it all had to call it "the North Pole Express" ....

Wednesday, December 22, 2004

German spam funnier than meafloaf

I received spam from some German source proposing some sillystuffen, and I replied:

"No sprechenzie Deutsch and no wanten zie spam" -- then I got this response, as translated by Altavista:

well you: -) now are you astonished which? times a Mail of me, well that has also an important reason! I chatten very gladly with you is only unfortunately my PC in the bucket and I so gladly still many more from you to know would learn have you desire which we a little at the mobile phone chatten? 0163/3299152 do not come is no frog I bite also,I genuinly much would be pleased! To then Anja

The bit about biting the frog has me especially concerned. Should the ASPCA be alerted? Or the French Anti-Defamation League?

Someone please call Anja and tell her to please pursue other continents....

Tuesday, December 14, 2004

My 2001 proposal for a digital Library of Congress comes closer to fruition

In 2001 I proposed that the nation should expend resources to digitize the holdings of the Library of Congress. Not a sample of materials on a single subject -- e.g. the American Memory project -- but every page image of every book in the collection -- tens of millions of books, full-text, full image -- terabytes of data.

Roy Tenant, a leading digital libraries expert, and I debated the concept remotely, with him at the Internet Librarian conference in California, and me teledebating from halfway across the world, from the First Monday conference in Maastricht, the Netherlands.

My main argument in 2001 -- and today -- is simple: disk is so cheap, and digital imaging technology is so mature, that you can

conceive of capturing every page of every book or journal in your collection at low cost.

Roy and I reprised this debate the next spring at the Computers in Libraries conference in Washington, D.C. Several hundred people listened to our debate. One of Roy's best lines was "but the Library of Congress contains a lot of junk!" My response: it's cheaper to digitize everything in the collection than to spend staff time deciding what is useful, and what's junk.

You have to understand that Roy is someone who actually creates digital libraries, and I am but a dreamer. So what the hell do I know? Still, I claim: You were right in a way, Roy. Large research libraries contain a lot of junk. But we retain the junk, and the real estate, because we don't know what is junk and what is useful.

A number of people came up to me afterward, including librarians at the Library of Congress, and said I'd put forth a grand vision. Sadly, LC did not pick up the mantle. Happily, three years later, Google has.

Here's a link to my presentation from 2001/2002:

The Digital Library of Congress

And here's Roy's side of the story from our debate in 2002. I must confess, he may still win the argument: this requires a huge amount of human effort, and a huge leap of technological faith. In researching my side of the debate, I found the earlier calculations of Michael Lesk invaluable. He understood the incredibly shrinking cost of disk before most people did.

Until today, every library digitization project was small-scale. No one took on an all-library project; they objected:

  • We can't afford to digitize the entire collection
  • It will take too long
  • We can't overcome copyright concerns

Google to the rescue. It takes a company with the vision, the technological prowess, and the capital that Google possesses, to make this happen.

One of the points I tried to make in my debate with Roy is that it would take a large-scale project, something on the scale of Kennedy proposing that the United States send a mission to the moon, in order to move digital libraries forward. When JFK proposed we send a man to the moon "before this decade is out" no one knew how the hell to accomplish that goal.

I believe this project is every bit as important as the Apollo moon program -- and this time, it's privately financed.

I believe today's news is cataclysmic, certainly in the halls of academe, if not a red letter day in world history.

Monday, December 13, 2004

University of Michigan announces 7 million volume digitization deal with Google

You heard it here first, received un-embargoed from secret source:

>The University today is announcing a groundbreaking
>partnership with Google that will digitize the entire seven
>million volumes in the U-M library and make them accessible
>via a simple Google search.
>This project puts the University at the leading edge of a
>movement that will transform access to knowledge. Anyone who
>has Internet access, anywhere in the world, will be able to
>search our entire library, without limitations of geography,
>time or expense. It is an endeavor that carries remarkable
>implications for our institution; as a great public research
>university there is nothing we care about more deeply than the
>creation and sharing of knowledge.
>The project will make it possible for a user to locate and
>read the full text of works that are out of copyright, and to
>find snippets of text for copyrighted material, along with
>information about where a work can be found.
>Google will begin placing digitized volumes online in
>mid-2005, beginning with materials in Buhr. The technology is
>non-destructive, and rare books are excluded.
>As a product of this partnership, the University Library will
>receive and own a high quality digital copy of the materials
>digitized by Google, and it will be able to provide enhanced
>access for University patrons. The digitization at this scale
>is a massive undertaking that we simply could not have
>achieved on our own. The University will receive no financial
>Harvard University and the New York Public Library are
>announcing their own agreements with Google today, and more
>may participate in the future.
>In undertaking this project, we understand and respect the
>copyright issues involved. As an institution we create, use,
>and distribute all sorts of copyrighted works, and we care
>deeply about copyright issues from all aspects.
>This project is consistent with the very purpose of copyright
>law as reflected in the U.S. Constitution, to promote the
>advancement and dissemination of knowledge.
>For more information about the project, go to

Sunday, December 12, 2004

A visit to the saddest town on Earth

Visited a nice little village today. I betcha the folks that live here haven't been happy since November 2:

Saturday, December 11, 2004

Google Suggests: Google learns its ABCs

New on the Google horizon this week: Google Suggests.

As you type, Google auto-completes your search with words or phrases that people search for the most. It offers a drop down that you can choose from; or choose to ignore the suggestions, and just keep typing.

A simple and powerful idea. What amazes me is how fast it is, and how little additional memory it seems to require for your browser. Clearly they've done some thoughtful engineering -- gee, what a surprise, the Google team writing good code!

Folks on Slashdot have examined the JavaScript code and report that it's doing quick little transactions back to the mother ship as you type. I'm betting they repurposed concept and code from Gmail, which does very clever auto-completes as you enter a person's e-mail address, and which carries on a conversation with frequently during your session.

Some folks were a tad confused when I showed this to them. Note that:

-- The suggestions Google gives you represent the most popular user searches Google has for the combination of letters that you type, not the most-linked-to sites.

-- The number of hits just represents the number of matching items in Google's index; it is not PageRank or a measure of importance. For instance, (sic, with the .com) has one hit, even though it's a highly important site.

My colleague Trevor Barnes came up with this formulation, a 26-part snapshot of what the world seeks in 2004.9:

The ABC's of Google

A is for Amazon
B is for Best Buy
C is for CNN
D is for Dictionary
E is for Ebay
F is for Firefox
G is for Games
H is for Hotmail
I is for Ikea
J is for Jokes
K is for Kazaa
L is for Lyrics
M is for Mapquest
N is for News
O is for Online dictionary
P is for Paris Hilton
Q is for Quotes
R is for Recipes
S is for Spybot
T is for Tara Reid
U is for Ups
V is for Verizon
W is for Weather
X is for Xbox
Y is for Yahoo
Z is for Zip Codes

Neil Young once wrote a song called "Rust Never Sleeps," named after an advertising slogan. Clearly, Google engineers never sleep.

Try it yourself:

Saturday, December 04, 2004

Yahoo Search and MSN Search clobber Google Search

Rumor on the street is that Google has improved its crawling coverage and frequency -- no longer on the 30 day rotation. Recent personal experience says Yahoo and MSN crawl Google's own spaces better than Google does.

I checked to see if Google Search had indexed an article I wrote on Wigblog days ago, a suggestion that Apple should sell Ipods pre-loaded with songs purchased from their Itunes service. Sadly, days after the posting, they had not:

So I checked to see if MSN Search had indexed the content. They had:

So I checked to see if Yahoo Search had indexed the content. They had:

So here's the deal: Google bought Pyra Labs over a year ago, and pundits predicted they'd integrate searching of Blogger content and thus build a huge competitive advantage. The truth is, Google does a poor job of indexing the Blogger sites they host, and their two main competitors index that same content far more effectively.

Score: Yahoo and MSN Search 1, Google zero.

Friday, December 03, 2004

Search logs as prediction

Search log analysis rocks. It tells a Webmaster what the customers want. Folks who know me well understand that I live in the search logs. I want to understand what people want to find, and the search logs are how I discern.

Lou Rosenfeld is in London, and in talking with a Financial Post person, Lou learns that the FP is much more pro-active than most log analyzers ever will be. Folks at the FP figure (UK pronunciation probably "figger") that if someone is searching for a company name (or the like) a lot suddenly, then maybe there's a story there.

Cool use of log analysis as crystal ball.

Using Search Log Analysis to Predict the Future