Articles Weblogs Books School Short Cuts Podcasts  
Peer Review - Richard Koman

Peer Review: How We Got From P2P to Hailstorm


It's time to start thinking of P2P in the context of Web Services. At O'Reilly and, we've always said that the thing that was interesting about P2P was not only that Napster users could screw the record companies (although that is interesting), but also that the Napster architecture, the distributed computing architecture, the instant messaging architecture, point the way to new ways of architecting the Internet.

One of the key advantages of these P2P systems is a reversal of the slashdot effect or the Star Wars trailer effect, in which the more popular content is, the less available is. P2P offers many, many redistributors of popular content and massive redundancy of content. In other words, the serving of content moves from the Center to the Edge. Same deal in distributed computing and most other forms of P2P. Computing resources move from centralized, expensive-to-build-and-maintain systems, to edge systems owned and maintained by someone else.

This is all old news. What is interesting to me these days, especially in the context of instant messaging, is the fact that P2P systems create enabling technology on which Web Services can build. If everyone is running an IM client, then you know a lot about when individuals are online and when they're not, and who they are. From there, it's not much of a leap to realize that you could start knowing quite a lot about what these individuals' preferences are, what kind of services they would like, and what they would pay for.

If you move services out to the Edge, where people really are, rather than insisting that users come into the Center (that's what cookies and other attempts to establish identity on a web site are about), you start mixing up notions of services with P2P in very interesting ways.

Get Peer Review Direct!

Sign up for the OpenP2P newsletter and read this column in email.

That's why we're so interested in Microsoft's work on Hailstorm, Passport, and .NET, and why instant messaging is moving to center stage in that scenario. Messaging and P2P are a fundamental building block on which peer-oriented Web Services can be built.

With that in mind take a look at these articles: