Am I FoRKing old bits again?
Dan Gillmor: Peer-to-peer is the way the Web was meant to work
BY DAN GILLMOR
Mercury News Technology Columnist
News and views, culled and edited from my online column, eJournal
P-to-P TO YOU AND ME: When people from Intel, Microsoft, and Sun
Microsystems start agreeing with some of the most fervent believers in open
computing standards, your immediate reaction should be puzzlement or
outright skepticism. But when the topic of agreement is peer-to-peer
technology, the concurrence makes perfect sense.
Peer to peer, or P-to-P for short, is still mostly floating around the
edges of technology's mainstream. But the buzz was real, and so was the
obvious progress and excitement, among the hundreds of tech folks who've
been thrashing out some of the possibilities this week at technology
publisher O'Reilly Associates' P-to-P conference in San Francisco.
The major promise of P-to-P lies in some relatively simple notions.
Collectively and individually, we're using only a small part of the value
of the devices at the edge of networks -- personal computers, for the most
part. Harnessing their power, which includes everything from processing
cycles to users' creativity, will make the Internet universe vastly more
valuable for everyone.
P-to-P is being defined broadly, perhaps too much so. But central to my own
view on it is the idea that every client -- that is, every PC and other
device connected to the Net -- should also be a server, dishing out
information to other devices.
Lots of people are working on this notion, which is nothing more than the
way the Web was supposed to work before corporate interests turned it into
a mostly read-only medium. The technical difficulties will be profound and
difficult to overcome, but as several people have noted, P-to-P is
ultimately the only workable computing architecture in a world where
billions of devices are connected to networks, because servers as we know
them today simply won't be able to handle such a large universe.
Much of the discussion was how to deal with some of the fundamental
infrastructure problems P-to-P will present. Bill Joy, Sun's chief
scientist and one of the Net's great contributors over the years, said his
company and some allies are working on some software called ``Jxta,''
pronounced juxta, a layer of code designed to be part of P-to-P's
fundamental plumbing. Precisely what Jxta will do wasn't clear, so we'll
have to take this on faith -- something that always makes me nervous when
dealing with tech companies.
Meanwhile, lots of companies you've never heard of are pushing hard to make
their presence felt, if not known yet, in the P-to-P arena. One is called
KnowNow (www.knownow.com), based in Menlo Park, and this company could be
onto something big.
Normally, a browser sends a request for information, which is delivered by
the server. The connection ends.
KnowNow holds the connection open. Then it adds some other software code
and, voila, you have a mini-server inside the browser. You're not
necessarily clogging the data pipes, but you are pretending, in effect,
that you're downloading a long document, slowly, while the browser keeps
communicating with the server.
The application potential here seems almost limitless -- real-time, two-way
communications with sophistication. There are also some security issues,
such as the possibility of compromising corporate firewalls. If this works
as described, however, it sounds valuable
This archive was generated by hypermail 2b29 : Fri Apr 27 2001 - 23:17:47 PDT