Atom Feed
December 12, 2002
Bryan Field-Elliot blogs a response to my puzzlement about Google’s page ranking. Interesting nitty-gritty info.
It’s well-known that Google prefers data other than what’s in the meta tags to decide on Page Rank, but I’m still surprised that the meta tags aren’t indexed at all.
Glenn Fleishman has an article in the NY Times today (registration required) that includes a map of the hotspots in NYC. Unsurprisingly, it maps to the racial and economic divisions of the city: “92 percent of network nodes were below 96th Street.”
Glenn mentioned this article in the hallways at Supernova a couple of days ago. Bob Frankston‘s reaction (and I hope I represent it fairly) was that the problem is really one of educating people about the benefits of getting connected: all but the very poorest in NYC have televisions, and if you can afford a TV you could have afforded a connected PC.
It does seem that connectedness would spread further down the economic pyramid if its benefits were clearer. But I think very few people, except for the upper reaches of geekdom, view TV and Internet as competitive technologies. Having a TV is close to a requirement for participating in our culture. As long as it’s an Either/Or, the digital divide in NYC will be real.
Mark Dionne emails me an interesting puzzle he’s discovered:
I just determined that Google does NOT index according to HTML “meta name=”keywords”” markup. I have a word in that markup in a page of mine that does not appear in the body of the page. If I hit the page with a search and then add the word, the page is no longer found.
In other words, let’s say I have a page about cats that puts “Siamese” into the meta tag for keywords but doesn’t otherwise contain the word “Siamese.” If the page contains the phrase “Cats love rubber baby bumpers,” a search on that phrase pulls up the page, but a search on “Cats love rubber baby bumpers” Siamese” does not.
Do we have any confirming evidence that Mark hasn’t simply run into an anomaly? Does Google discard keywords because they are often used to torque the system? Is this yet another argument against relying on humans explicitly tagging content? Is this a death knell for the Semantic Web? Will Ashley discover that Grayson is not only her lover but is also her transsexual twin sister? Tune in for “As the Web Spins”…
December 11, 2002
Red-eye last night back from Supernova. Glenn Fleishman gave me a ride to the airport, but after that I was plunged into the pleasure-deprivation mindset of the airlines. No food, no pillows, no blankets and “attendants” who become brusque and defensive if you ask for the same even if you use a tone as falsely pleasant as their own.
Yeah, I’m grumpy.
Thank goodness for Dramamine (or its brandless equivalent). A pharmacist once recommended it to me as an over-the-counter soporific. It works for me and leaves not much of a hangover. Except for the grumpiness.
December 10, 2002
Kevin is summing up. He’s asking us what we haven’t addressed enough.
His answer is that we haven’t talked enough about government and about digital identity. Here are some of our answers: [I didn’t get everyone’s name because the ideas were pumped out rapidly.]
Marc: The mechanisms for hooking together decentralized stuff. (This is more techie than I’d care about.)
The economic reasons for decentralization
Doc: Categorizing infrastructure, e.g., commons vs commercial infrastructure.
Dave Sifry: Business models for edge stuff
Organzational and legal models
SIP, IM and super directories
Grid computing
Gary Bolles: Inside the firewall where decentralization looks bad
Fair Use and DRM
Consumer electronics and non-PC computing
Phil Wolff: In the next Supernova, which will be in DC, we should focus on explaining this stuff to policy makers.
Excellent conference. Now I get to ride to the airport with Glenn Fleishman.
I seem to have suggested that some conference be held at my house and AKMA has decided that it’d be a swell idea.
I am so fried from the Supernova Conference (which my body is currently attending) that this is immensely unappealing to me. But, I don’t want to go back on my offer. So, sure, let’s have a conference at my house.
Unfortunately, I’ll be out of town that day.
I zoned out for the first 30 minutes of this panel. Some distuurbing email came in that I wanted to deal with. And I’m pretty much at my limit for sitting still and listening to smart people. Will someone please turn on “Friends” or “Buffy”? Thanks.
Fortunately, I was able to turn to Doc’s weblog where he’s entering something like a smart transcript. So, I’m listening to the panel while reading the blogging about it…
No wonder my head hurts.
David Isenberg is talking about what’s happening in the world of telecommunications. He’s just back (two nights ago) from several weeks in Japan, Australian and New Zealand so he’s full of wide-eyed news of a world where broadband flows like milk and is as sweet as honey. Now he’s talking about a vanilla 802.11b system that provides better quality sound than “real” telephones.
“What HTTP did for documents, SIP (Session Initiation Protocol) will do for communications.”
“The best network is a stupid network.” That is, the best network provides nothing but connection. The services are provided by applications running on the network. (David and I wrote about this at The Paradox of the Best Network. All content came from David.) “Each of us in 2-3 years can have the bandwidth of a telephone company for a few thousand bucks. But the telephone companies believe in scarcity and are forcing it on us.”
The End to End principle, which is basically the same as The Stupid Network, says that you should keep the network simple because that preserves your options for innovating on its edges. Phone companies like to add value to their networks, for competitive reasons, which makes their networks smart. That’s fine for telephone calls, but the Internet is not specifically for phone calls or for anything else.
The Internet, says David, makes telephony just another application. This is a problem for the business model of the telephone company, for the telcos know about stringing wires and optimizing telephone switches (an expense) and make money selling the application for monthly revenues. In a stupid network, the application provides revenue. But, asks David, where’s the business model for the physical layer of the Internet? In short, who builds and runs the new network? Not the telephone companies because it’s too foreign from their business model. The cable companies are unlikely because they don’t want to enable the wholesale generation and sharing of entertainment that competes with the entertainment they’re selling. Municipalities? Sometimes. Utilities? Maybe. New companies? Customers?
There are, he notes, important political implications and obstacles. But he’s out of time.
Smart presentation on the virtues of stupid networks. But, then, I’m partial to David. I’ve learned a lot from him. (His newsletter is free and always excellent.)
Kevin is interviewing Sergey Brin.
Google Groups (the Usenet search capability) came about because an engineer at Google, Michael Schmidt, dreamed of it. Same for the Google API: An engineer there was enthusiastic about doing it. Cool.
What’s the business model for the API? There isn’t one. Right now you’re limited to 1000 hits a day, and maybe someday they’ll charge. But they did it because it was interesting and fun. His example of an API-based app is a plagiarism checker.
Marc: Will Google support XML structures so we can get better searches. E.g., searching for images.
Sergey: Rather than making humans write better, we should work on having computers understand better what humans are writing for one another.
Dave: Do you envision Google ever becoming decentralized? A Google for the desktop?
Sergey: I’d like to be able to search my desktop from anywhere. But we haven’t written client software. Our technical competency is running large server farms.
Sergey just did a show of hands to discover that Google is indexing the blogs from most people in the room but isn’t updating as frequently as we’d like. (We are a spoiled bunch of puppies!)
[I went to lunch before posting this and somehow lost most of it. Odd. Fortunately, you have another 28 blogs covering this to choose from.]
I asked if Google has considered optimizing itself to support its use as a DNS extender (in Dan Gillmor’s sense: since bobsmith.com is owned by only one Bob Smith, I’ll use Google when I want to find my old pal Bob Smith). Sergey didn’t really answer this question, though, preferring to answer something about the difficulty of handling complex searches.
Cory gots to ask the last question: Do you Google guys ever think about the way the world is reshaping itself around Google? …Good to end on a question that’s actually a thank you.
This should be great. The panel is Meg Hourihan, Dave Winer and Nick Denton. (In fact, it was really good. But because it was conversational, the following account is fragmentary.)
Dave is blogging the panel as he’s on it. Call it Escherblogging.
Dave says he did XML-RPC and Soap to enable people to write to the Web as easily as they write with their word processor. Weblogs and Web services are all about the human needs they serve. (Write on, Dave!)
Nick sees weblogs as a way to produce media more cheaply. “Weblogs are one way to produce media properties that distribute the editorial processs…”
Meg cares about weblogs because it enables people to connect.
Meg defines a blog as something that’s updated daily, presented in reverse chronological order, time-stamped, etc. But wouldn’t that make the AP headlines page a blog? I ask if blogging is not just a technology but also a tone of voice — a new type of rhetoric, one that’s distributed over the Web.
Meg on whether rightwing yahoos are going to take over weblogging: “It’s not like talk radio. There’s unlimited bandwidth. Anyone can have a weblog.”
What percentage of people online will have a weblog? Dave: Everyone.
Nick: Ten times as many people writing in public than in the world of print. And that’s a big deal.
Dave: In 10 years, every member of Congress will have a weblog.
Mitch Ratcliffe: There’s no point in arguing in what is or is not called a weblog. What matters is that more and more people are communicating in their own voice.
Cory:”Blogging allows people to write who couldn’t write before… It’s not the usual suspects.” What makes it easier is the informality of blogs as well as the ese of use of the technology. For example, EFF set up a blog so they could get past the barrier of having to have every pronouncement be official and thus have to go through all the official poicies and controls.
Final question: What do you want from blogs in the next few years?
Meg: Push. She wants to have the content come to her.
Dave: What does the audience want? Doc: Search. Me: Blogthreads. Marc Cantor: We’re doing a “topic server” that will enable blog threads.
Nick: Go to one page and see the buzziest items of the hour, recommended by my social network.
Joho the Blog uses WordPress blogging software.
Thank you, WordPress!
