Slashdot Mirror
In Google We Trust
firstadopter.com writes "The New York Times (registration needed) writes about how far Google has penetrated our culture (soul sucking "Free" registration required) in the last six years with the pros and cons of its success. It's amazing to think 200 million searches are done on the search engine each day on an index of 6 billion pages."
Yep, while listening to Michael Myers commentary track of the Goldmember DVD, I heard him call Michael Caine a "veritable Google of the entertainment business." Thus, we are stuck with the word google as synonymous with search or knowledge base, whether Google likes it or not.
-- SKYKING, SKYKING, DO NOT ANSWER.
In Searching We Trust
By DAVID HOCHMAN
Published: March 14, 2004
BEN SILVERMAN is what you might call a Google obsessive. A producer and a former talent agent best known for bringing "Who Wants to Be a Millionaire" to American television, Mr. Silverman Googles people he is lunching with. He Googles for breaking news, restaurant reviews and obscure song lyrics. He Googles prospective reality-show contestants to make sure they don't have naked pictures floating around the Web. And, like every self-respecting Hollywood player, he Googles himself. Competitively.
"Guys all over town are on the phone saying, `I bet I can get more Google hits than you.' " he said recently. "It's become this ridiculous new power game."
It's more like the new kabbalah. With an estimated 200 million searches logged daily, Google, the most popular Internet search engine, "has a near-religious quality in the minds of many users," said Joseph Janes, an associate professor at the University of Washington in Seattle who taught a graduate seminar on Google this semester. "A few years ago, you would have talked to a trusted friend about arthritis or where to send your kids to college or where to go on vacation. Now we turn to Google."
The Web site that has become a verb is many things to many people, and to some, perhaps too much: a dictionary, a detective service, a matchmaker, a recipe generator, an ego massager, a spiffy new add-on for the brain. Behind the rainbow logo, Google is changing culture and consciousness. Or maybe not ? maybe it's the world's biggest time-waster, a vacuous rabbit hole where, in January, 60 million Americans, according to Nielsen/Net Ratings, foraged for long-lost prom dates and the theme from "Doogie Howser, M.D."
"In one sense, with Google, everything is knowable now," said Esther Dyson, who publishes Release 1.0, a technology-industry newsletter. "We were much more passive about information in the past. We would go to the library or the phone book, and if it wasn't there, we didn't worry about it. Now, people can't as easily drift from your life. We can't pretend to be ignorant." But the flood of unedited information, she said, demands that users sharpen critical thinking skills, to filter the results. "Google," she said, "forces us to ask, `What do we really want to know?' "
Google delivers information that can radically alter one's self-perception. About a quarter of "vanity" searchers ? those who search for their own names ? say they are surprised by how much information they find about themselves, according to a survey by the Pew Internet Project.
Sometimes, they're really surprised. When Orey Steinmann, 17, of Los Angeles, entered his unusual name on Google's query line, he discovered that he was listed on a Canadian Web site for missing children and told a teacher. After an investigation, county officials took him into protective custody last month and federal marshals arrested his mother, Gisele Marie Goudreault. She has been charged in Canada with parental abduction, said Barbara Masterson, an assistant United States attorney in Los Angeles. Canadian authorities are seeking Ms. Goudreault's extradition, and Orey is deciding whether to contact the father he never knew.
Then there are the Google miracle stories. The morning after five left-handed electric guitars owned by Robert McLaughlin were stolen from a storage room at his San Diego apartment complex last year, he searched Google's image library for guitar photos to use on a reward poster. Instead, he found the stolen goods. "The thief was selling them in a live auction," he said. "In the past, my report would have gotten lost in a mountain of paperwork. Because of Google, the cops recovered four of the five guitars that week."
While some compare Google's reservoir of six billion documents to the ancient library at Alexandria, it often feels like the shallowest ocean on earth. "Google can be useful as a starting point to research or for superficial inquests," said James H. Billington, the Librarian of Congress. "But far too often, it is a gateway to illiterate chatter, propaganda and blasts of unintelligible material."
Fun new one to try: Mooter
For developers, the Google SOAP API is great. I used it a year and a half ago for a demo system that answered "who" and "where" questions posed in natural language. You need to ask for a license key that allows 1000 SOAP based calls a day. In addition to searching, you can also use the Google spelling corrector with this API.
Amazon also provides a SOAP (and REST) API.
-Mark
I don't think [b] has anything to do with the page-rank score, I was just using it as an example.
"Symantically correct" html means the tags have meaning. [b] (bold) doesn't *mean* anything. Neither does [i] (italic) or [font]. The preferred tags to use are [strong], [em] (emphasis), and [h1-6]. This idea is that HTML should describe content, and stylesheets should determine how the content looks.
If you surround something with [b] tags, you're coupling the content and the presentation. It's better practice to surround content with [strong] tags and then define how [strong] looks via a stylesheet.
200 million searches a day, eh? Being a performance geek, I am driven to estimate the implications of that load.. Please feel free to augment and correct..
200M searches/day = 8.33M/hour = 138888/min =
*** Google averages 2415 searches / second ***
Average page size = 5,563 bytes (a search for "apple", hey I RTFA)
Assume outbound bandwidth requirement of 6000 bytes/search with some overhead.
2415/sec * 6000 bytes/search =
*** 13.88 MB/sec avg or 1200 GB/day bandwidth requirement (OUTBOUND ONLY) ***
CPU.. 2415 searches/second.. Determine required aggregate CPU capacity using various assumed values for 'CPU per search':
0.25 CPU sec/search = 603 CPU seconds required for each wall second
0.5 CPU sec/search = 1207
1.0 CPU sec/search = 2415
2.0 CPU sec/search = 4830
4.0 CPU sec/search = 9660
8.0 CPU sec/search = 19320
Assume they only run the search boxes at 50-80% util and tweak estimates accordingly. Also, the burstiness inherent in the internet will greatly impact these requirements (assume at least +30% for the second to second variations as well as the hourly variations).