IntegrityP2P's Interview with Rodi Developer Larytet
By Vivacious

IP2P R&D Team
April 23, 2005

Have you been following filesharing news? If you have you're aware that the music and movie industries are hard at work trying to stamp out innovation and protect their already fatted wallets. Open source projects, like VLC Player are being threatened by European lawmakers. It would seem a dark time for peer to peer application development. One developer is undeterred.

Rodi developer Larytet decided to create an open source, fully decentralized filesharing application and network with the ability to protect content providers.

Recently the IP2P Research & Development Team has taken a good look at newcomer p2p application Rodi and asks Larytet for his views on Rodi and his aspirations for the future of peer to peer.
__________________________________________________ _____
Q:Can you give us an overview of what Rodi's
features are and what it can do?

At the current phase this is not much - users can search
the network using a filename or filehash (MD5) as a key,
download and upload/share files, and
run bouncer to proxy traffic for others. This is
approximately 20% of initial functional requirements.

Main market for Rodi is LANs like corporate networks
or campuses. Satellite internet where TCP underperforms
is another.

Q: What inspired you to make Rodi?

Probably the last drop was closing most popular
torrent servers. I started to understand that this is
probably much more serious than just MP3s or poor
quality video clips. I decided to create a fully
searchable and decentralized network, where publishers
or content providers could stream data without
exposing IP addresses.

Q: Where do you see P2P at in ten years, and what
role do you think Rodi will play then?

Does your question imply, "What is going to happen with
distribution of copyright works" ?

I expect that in the next 5 years authorities in most
countries will require ISPs to provide realtime
access to the full logs. Statefull firewalls are going
to be mandatory if the ISP wants to keep their license. It will
greatly influence both P2P and content providers. I
think that the process is going to be bidirectional. Content owners
will reduce prices and distribute many
more titles in attempt to monetize "tails" -the 60% of
customers who are not satisfied by pop music. On the
other hand law enforcement is going to be
significantly more effective than it is today.

If you ask about the legal applications of P2P, like Skype
and Linux distros over BT it depends on how tolerant
ISPs are going to be to the "parasitic" traffic. So
far ongoing events do not provide many reasons for
optimism. And if ISPs get access to the content at
reasonable prices (think AOL/TimeWarner) there isn't
any business reason to run P2P based distribution
networks. Video server can stream the data as well.
Another direction is grid networks. There are some
other areas, but all of them are relatively small
niches and I think out of scope of the question.

I guess that Rodi will not survive 10 years, but I
hope that some ideas from Rodi will.

Bottom line - traffic analyzers can be very effective.
Enforcement of this or that policy is technologically
possible today and getting cheaper and easier every
minute. There isn't much money discovered in P2P
applications, bu
t there is plenty of money invested
in the firewalls and NATs because of corporate
networks.

I would like to see that free "open source" artistic works create
competition for the existing industry. In my view this is one of
the most promising areas of using P2P.

Q: How does Rodi differ from other P2P apps?

This is probably the most decentralized P2P network
created so far. Rodi can run without super nodes,
GWebcashes, message boards, index servers, trackers.
two peers can find each other even from behind
firewalls and NATs.

To my best knowledge this is the first P2P network
functional requirements of which include things like
faked RTP packets to circumvent traffic analyzers and
firewalls.

This is the first attempt to create a P2P network which is
searchable by content and effectively creates a fully
decentralized search engine. Imagine that instead of
Yahoo/Google/MSN everybody can run his/her own search
engine and produce results according to the customized
keywords tables. The Isaac Asimov fanclub can run their
own search engine, and the Vietnam Vets of America their own.

Q: What kind of protections does Rodi offer to its
users?

Packet bouncing and the possibility to spoof IP source
address.

Rodi bouncers behave exactly like any proxy in other
anonymous networks - forwarding packets to the
destination.

IP spoofing is not possible for many, maybe not for most
of the broadband users, but for those who can spoof IP
source, Rodi is probably the only alternative.

Q: Rodi features "NAT Penetration", how does that
feature work, and who can benefit by using it?

In the current version Rodi client creates malformed
DNS packet and sends it to port 53 (DNS). It means
that the Rodi client behind a firewall can send data only to
Rodi clients using port 53. This is not the only
limitation, but the most important one.

It works because apparently some firewalls do not
implement any analyzing of DNS packets.

In the future I expect that I'll have to implement
spoofing of RTP packets and HTTPS.

Everybody sitting behind a firewall (University,
business) can run a 300K applet directly from a WEB site (Try Rodi)
or a USB disk or a floppy and gain access to the
network. Part of the functional requirements is HTTP
tunnelling. The moment it's implemented every Rodi peer
is going to be an HTTP proxy. It will create much more
work for companies like Websense. ...

Read entire article on IntegrityP2P.com.