From: Adam Rifkin -4K (adam@XeNT.ics.uci.edu)
Date: Mon Jul 24 2000 - 16:41:40 PDT
This statement sounds like something Rohit would say...
| "There are about 100 uses for peer-to-peer," says Graham Spencer, a
| former official of Excite At Home Inc., and an InfraSearch investor.
| "So far, we've thought of maybe two of them."
Anyone have any ideas for the other 98? Looks like we still have a few
weeks' window to jump on this train... but I'm not sure I'd go so far
as to call this a once-in-a-decade phenomenon...
| Boosters -- including Marc Andreessen, co-founder of Netscape
| Communications Corp. -- say the technology is one of those
| once-in-a-decade ideas that will change the face of computing.
Methinks MarcA is being a tad melodramatic... can you imagine having a
hundred different Napster-like applications running on your desktop at
the same time? Me neither. Napster alone is clogging up my pipes, and
[Napster+ICQ+Browser] is slurping up all my free available attention.
Let's see who gets a mention in the article: InfraSearch (which in the
two weeks since this article has changed its name to GoneSilent as Mark
Kuharich noted), Kalepa... and hey, there's Centrata again and now they
have KP behind 'em. Can I call them "Firedrop II"? :) And somehow KP
investment Google managed to sneak their way into this story even though
they're not a P2P company at all.
No mention of AppleSoup or Popular Power. Maybe they'll make the
> Venture Capital Loves 'P-to-P,' The Latest Technology Fad
> By LEE GOMES and LISA BRANSTEN
> Staff Reporters of THE WALL STREET JOURNAL
> July 5, 2000
> Despite the dot-com downturn, Silicon Valley is embarking on one of
> its favorite pastimes: investing in a full-blown technology fad.
> This one, called "peer-to-peer" computing, is inspired by the
> controversial music-sharing service Napster Inc. Instead of Internet
> information being held in a few central locations, peer-to-peer
> computing makes it theoretically possible to unlock the files and
> data residing on every personal computer connected to the Internet.
> Boosters -- including Marc Andreessen, co-founder of Netscape
> Communications Corp. -- say the technology is one of those
> once-in-a-decade ideas that will change the face of computing.
> Peer-to-peer, for example, could be used to create online auctions
> without any middleman, or Internet searches that find up-to-date
> information from many sources besides Web pages.
> Inspired by such possibilities, Mr. Andreessen and other
> deep-pocketed investors are starting to write big checks to fund
> companies pursuing the new approach.
> "It's a hot space," says Josh Becker, an associate at Redpoint
> Ventures, a Menlo Park, Calif., venture-capital firm. Business plans
> are coming in, he says, mostly from "really young hackers. The next
> month will be key. That's when a lot of the winners will be
> Skeptics, and there are many, say the accumulating froth over
> peer-to-peer reminds them of other big ideas that came and went
> without ever living up to overheated predictions, like "push
> computing," which in 1996 was set to replace traditional Web browsing
> as the way people would access the Internet.
> In peer-to-peer computing, two computers -- such as those belonging
> to two home Internet users -- connect to each other directly, without
> the help of the massive central machines that traditionally run big
> Web sites like Yahoo! Inc. Napster, which spawned interest in the
> technology, is considered peer-to-peer because users of the program,
> once they learn about each other through Napster's computer, end up
> trading MP3 music files directly with each other, without Napster
> being involved. An even purer version of peer-to-peer is Gnutella, a
> Napster-like program that dispenses with the central database
> Napster and Gnutella are most commonly used to look for bootleg MP3
> music files. The same approach is also expected to be used for
> sharing other types of files, such as computerized copies of books
> and movies.
> Most of the investors interested in peer-to-peer these days aren't
> thinking about Napster-like applications, but rather, entirely new
> uses -- especially for new kinds of search engines to supplement the
> familiar ones offered by Yahoo and its rivals. That concept helped
> spawn InfraSearch Inc., the company Mr. Andreessen is investing in.
> The San Francisco start-up is headed by two recent graduates of the
> University of California, Berkeley: Gene Kan, 23 years old, and
> Yaroslav Faybishenko, 22. While they had little in the way of a
> formal business plan, the two men have raised nearly $2 million by
> capitalizing on investor interest in what seems like the next big
> Mr. Kan and some friends had started a Web site about Gnutella, and
> he emerged as something of a spokesman for the software, even though
> he had nothing to do with its creation. At the same time, Mr. Kan
> started thinking about other uses for Gnutella technology, especially
> for searches. He helped put together a demonstration that was a
> primer in peer-to-peer searching.
> It connected four computers: a calculator, a database full of
> pictures, a collection of news stories and a Web site that dispensed
> stock quotes. When a user entered in a search request, Gnutella
> passed the request to each computer, which would respond the best way
> it could.
> For example, typing "2 + 2" would get a sum from the calculator but
> nothing meaningful from the other machines. But typing "cat" might
> get a picture of a cat from the photo archives, recent news stories
> mentioning kittens from the news collection and a stock price for
> Caterpillar Inc., whose ticker symbol is CAT.
> Those results are much different from what happens in a traditional
> Web search. Those services regularly pull information from millions
> of Web sites and place it their own repository, a process called
> "crawling." Searching this centralized index gives results quickly,
> but it only includes information contained on Web sites and is only
> as current as the last time it crawled the Web.
> A peer-to-peer search, in theory, could be more current and reach
> non-Web sources, such as databases maintained by companies or
> individuals. Participants in a peer-to-peer network would decide
> what, exactly, they wanted to share on the network. An individual
> user, for example, might allow one of the folders on a hard drive to
> be scanned by network users, as happens with Napster. Or, a computer
> running a database could allow requests coming over the network to
> make direct queries of the database, and then report back the answer.
> Any peer-to-peer system would need to guarantee the privacy and
> security of its users, or it would probably not find many customers.
> The InfraSearch demo attracted interest from tech-savvy "angel"
> investors in Silicon Valley. Mr. Andreessen, who helped invent Web
> browser software before joining Netscape, was an obvious target.
> Hearing from a reporter that Mr. Andreessen was intrigued by
> Gnutella, but not knowing his e-mail address, Mr. Kan sent e-mail
> messages to Mr. Andreessen at the latter's new company, using every
> variation on Mr. Andreessen's first and last names until one of the
> messages got through.
> "He reminds me of me six years ago," said Mr. Andreessen of Mr. Kan.
> "This is a big, new idea the same way that the browser was a big, new
> idea, and there will be a lot of people going after it."
> Other companies are also jumping into the peer-to-peer space. Miko
> Matsumura, a former employee of Sun Microsystems Inc., helped found a
> Palo Alto, Calif., company called Kalepa Networks Inc. to connect
> different peer-to-peer networks. Numerous other companies, such as
> Centrata Inc. of Cambridge, Mass., are exploring variations on the
> theme of peer-to-peer computing. Shishir Mehrotra, chief executive of
> Centrata, said that his company has just closed an investment from
> Kleiner Perkins Caufield & Byers, the Menlo Park venture-capital firm
> and that partner Vinod Khosla would take a seat on the company's
> And Michael Tanne, entrepreneur in residence at Redpoint, says that
> the firm is working with at least two companies in the area whose
> very names haven't even been announced yet but are planning to
> announce their businesses in the next few weeks.
> Peer-to-peer fans see several ways the technology can unlock the
> resources of the Internet. A newfangled version of eBay Inc., for
> example, could allow people to use their own computers to list items
> they have for sale and complete transactions with people looking for
> bargains. Such a plan could avoid the need for a central service that
> takes sales commissions.
> But there are several technical hurdles. Peer-to-peer networks
> require each participating machine to pass along all network traffic
> to the next machine. Many doubt if home computers will ever be
> powerful enough to handle the untold millions of messages that would
> fly through a global network. And like all other search engines, a
> peer-to-peer system would still need to sort through its heap of
> results to find the exact information the user wanted.
> Mr. Kan says his colleagues have ideas about dealing with both
> problems. They also will have to design the system to make sure that
> new computer viruses can't be spread when home users' hard drives are
> accessed as part of a peer-to-peer network, though nothing like that
> has yet happened with Napster.
> Messrs. Kan and Matsumura say that they are still working on how they
> will actually make money with the technology. One candidate: selling
> peer-to-peer technology to big companies to tie all their computers
> into a single large database. But the two men are likely to face
> competition not only from upstart peer-to-peer companies, but also
> from traditional search engines. Those centralized approaches, says
> Sergey Brin, co-founder and president of Google Inc., a search site,
> are learning to offer some of the advantages of peer-to-peer, like
> more timely access to information, without all the associated
> technical burdens.
> Backers are undeterred. "There are about 100 uses for peer-to-peer,"
> says Graham Spencer, a former official of Excite At Home Inc., and an
> InfraSearch investor. "So far, we've thought of maybe two of them."
> Write to Lee Gomes at firstname.lastname@example.org or Lisa Bransten at
eNow: A million minds a minute. -- http://www.enow.com/
This archive was generated by hypermail 2b29 : Mon Jul 24 2000 - 16:42:07 PDT