Written by Can Erten and edited by Richard MacManus. This is the second in a 2-part series exploring the world of P2P on the Web. Part 1 was a general introduction to P2P, along with some real-world applications of P2P. Part 2 (this post) discusses future applications.

As we mentioned in Part 1, broadband speeds are ever increasing and so the demand for peer-to-peer networks is also increasing. However many things that could be accomplished by P2P networks are still in development or research. There is huge potential that at least some of the resulting applications will go mainstream, just as Napster did in the late 90's or Skype in the early part of this century. In Part 2 of our series, we look at some of these potential future applications for P2P on the Internet.

Search Engines

Starting in the late 90's, a search engine company called Google changed the way we search the internet. Their idea was to index the web and get the top results, using their now famous Page Rank algorithm. However nowadays, indexing the web accurately has become a huge and seemingly impossible job to complete. So P2P search engines could be the next solution - where every node (user) is a crawler itself.

In P2P search (a.k.a. distributed search), each individual connected to the network serves its local index as a source of search. Instead of having a central company and a central server, each participant of the network is a search repository. Since we are talking about web indexing and web searching, a user's internet cache might be their contribution to the search database. When they execute a query, firstly their local system is queried; than if the results are not satisfactory, the next peer is queried, and so on. The difficulty is the selection of good peers to provide satisfactory results. P2P search may well be a long shot, but one possible solution in this area to check out is the Minerva Project. It is described as follows:

"Each peer is considered autonomous and has its own local search engine with a crawler and a corresponding local index. Peers share their local indexes (or specific fragments of local indexes) by posting meta-information into the P2P network. This meta-information contains compact statistics and quality-of-service information, and effectively forms a global directory. However, this directory is implemented in a completely decentralized and largely self-organizing manner."

Video and Audio Casting

The impact of video and audio web sites on the Internet has been very large over the last couple of years - and will only increase. Therefore there has been talk of moving video streaming to P2P networks, to lessen the load on the Internet. A P2P approach for video streaming would be to hold a copy of a file in different parts of the world and serve it from multiple points to users.

The creators and entrepreneurs of Kazaa and Skype, Zennstrom and Friis, are working on such a project - called Joost, a.k.a. the Venice Project. It will be like a TV on demand service, but based on P2P where clients connect to the network and download TV programs. Joost also features social networking aspects - you can rate and discuss TV programs with other people. At the moment Joost does not provide a lot of channels, but the potential is there once more content is added. See also Read/WriteWeb's earlier review of Joost and other IPTV services.

Mobile P2P Applications

Many popular web applications have been ported to mobile platforms already. Likewise there is huge potential for P2P mobile applications, at least when wireless network enabled mobile phones become more popular. I think it will follow the same trend as for PC P2P applications - i.e. it will start from instant messaging, followed by file-sharing and IP telephony, then video and other media. Already Skype uses P2P for its VoIP, as we explained in Part 1. Also back in 2001 Swedish software maker Pocit Labs developed a mobile file sharing client called BlueTalk. It enabled file sharing over Bluetooth for up to 54 people - for example to trade files or play networked games.

A more recent example is PeerBox, reviewed earlier this year by ThinkMobile. PeerBox allows you to search and download music, videos and pictures; among other things.

E-commerce

Consumer to Consumer e-commerce is one of the most popular services on the Internet. A centralized trading platform (such as eBay) enables consumers to trade, buy or sell their goods. However in a centralized system, there is always a possibility of a failure - such as the server goes down or is busy. P2P enabled e-commerce can remove the centralized system and so lessen the possibility of failures. However there are many things that have to be implemented for a P2P system for e-commerce to work - it has to be secure, transactional and workflow-based to track different stages of the sales process. It also has to support detailed search, e-commerce advertisements and location awareness of the peers.

One early example perhaps is Tamago, a P2P marketplace that has been reviewed before on Read/WriteWeb.

Conclusion

In this two-part series, we've examined different types of P2P systems and their applications - past, present and future. We've covered just a few areas (check the comments on our previous post for other exciting applications).

Let us know some of the promising new applications of P2P that you've come across, which Read/WriteWeb could perhaps explore in future articles.

Image credit: MFA Design Technology