Anonymity and trust aside, there are surely other reasons that peer to
peer is useful. Efficiency, for instance - if peer to peer technology
enables us to usefully harness resources that were otherwise idle, then
that is a good thing. What is more, I'm not sure your arguments that a
centralized solution can provide all of the benefits of peer to peer are
necessarily valid. Can you really say that google is pure client/server
just because we access it through a web browser? The server side
implementation may well be using peer to peer techniques, although I
personally could not confirm whether or not it does.
The fact is that peer to peer techniques lead to desirable properties
such as failure tolerance and scalability. They are appropriate not only
for Internet-wide systems like Gnutella, but also for systems running on
private networks at places like Google. Infact, since security in peer
to peer environments is such a tricky problem, I think it is likely that
environments like this will provide the most immediate applications for
peer to peer research.
Matt
Tien Tuan Anh Dinh wrote:
Hi, (i posted this in another mailing list, but repost it here for
more discussion)
Let's look at the "traditional client-server model" in another way,
with Google as an example. Apparantly, Google employs a very
well-engineered distributed systems with thousands of machines all
over the world, but i think in the essence, it still follows a
client-server model. All searchs from a Web browser (client) goes
directly to a Google server.
We go on and on about how P2P systems allow decentralization of their
services, hence difficult to take down and can be immuned from DoS
attacks on their availability. Now look at google, i dont know how
many times it suffered from such attacks. None ??
What P2P services that google doesn't have ? Global searching ?
You're taking about the master here. Global storage system ? Google
has no problem in giving you 2-3Gb freely(emails, pictures and other
spaces). Sharing ? Google video, Picassa and even Google music.
So i come to think what the future of P2P would be like, especially
when the day that all file-sharing networks got shut down by RIAA. In
the user point of view, what advantage that P2P systems have over
Google (or other client-server modeled system with load of money).
Technically, is there any P2P service that cant be provided with equal
quality by a very rich company ?
Then, i stopped being depressed for one second, when looking at the
security issues. Many ways to define privacy, but for me, it consists
of anonymity and unlinkability (no correlation). You won't get
anonymity when using Google (search, GMail ...). Is there any truly
anonymous P2P systems out there ? At least there are undergoing
development for such systems. Correlation is really a pain in the *ss,
Google correlates all your search and your mail and store that on the
database, it knows you better than you know yourself. Once that
database leaks, or got seized by the police, your privacy is seriously
violated, and worse thing is that nothing you can do about it. P2P
behaves much better in this situation, as your activity (search) were
distributed among all other peers, and they won't be interested in
keeping your (log) information at all. It's also hard for the police
to track down your activity on the net; they'd have to look at other
peers' computers as well, and those peers will definitely not give in
easily.
In summary, my big question is, will there really be a bright future
for P2P ?
_______________________________________________
P2prg mailing list
[EMAIL PROTECTED]
https://www1.ietf.org/mailman/listinfo/p2prg
_______________________________________________
p2p-hackers mailing list
[email protected]
http://lists.zooko.com/mailman/listinfo/p2p-hackers
_______________________________________________
p2p-hackers mailing list
[email protected]
http://lists.zooko.com/mailman/listinfo/p2p-hackers
