Support
|
|
|
Help Save Reptile!
|
Navigation
|
Essentials
Installation
Developers
P2P (content distribution)
Search Infrastructure
Services
Proposals
Resources
|
|
What is Reptile?
|
Reptile is a P2P (peer to peer) application designed to locate
and filter the best news on the Internet. Reptile provides a
distributed and decentralized mechanism to search, cache,
subscribe, and publish news and other content. Reptile also
provides an infrastructure for increasing information
diversification and reducing censorship and bias.
Reptile is decentralized. No single point of failure should
deny a user from quality news. To this end, we also to 'bind'
across multiple network architectures. Reptile runs over the
conventional "web" (HTTP) but also runs over more modern and
distributed P2P architectures (JXTA).
Reptile is designed around a hybrid infrastructure which
supports the advantages of both client/server and P2P systems.
For example, one could run Reptile as a P2P system on a laptop.
One could also run Reptile as a client/server application from a
home computer and access it over SSL from a web browser from an
outside location. Reptile also supports rendezvous nodes
(supernodes) which bridge this functionality with the rest of
the world. For example a major website can run a search request
via a stable Reptile node running on a known host (AKA
openprivacy.org).
|
|
Architecture
|
Distributed searching
|
Reptile provides a distributed searching infrastructure.
Reptile nodes can use each other as search providers. They
can also search Reptile "supernodes" (nodes which are
available for long periods of time and have large databases)
for greater availability. Since this is based on a service
infrastructure. Third parties can develop and work with
Reptile services.
|
|
Distributed caching
|
Reptile nodes can exchange cached information which each
other. This is provided so that if a major website fails,
Reptile nodes can still obtain the information. This allows
provides a much greater level of scalability.
|
|
Distributed subscription and publication
|
Reptile users can publish information and other Reptile
users (and websites) can find this news and subscribe to it.
This also works in the both directions as Reptile users can
subscribe and publish to each other's channels.
|
|
Multiple network support
|
Reptile binds around multiple network infrastructures. The
convention web as well as P2P systems are supported. This
allows us to support the best of both systems and allows for
a fair playground for anything with an electronic heartbeat.
|
|
Reputation optimization
|
Reptile users can use
reputation
(provided by the Sierra framework) to help
filter out low quality (or irrelevant) information. Users
can create trust
Our reputation
system is still under development. The main goals are to
build a system which is highly scalable, allows each user to
define others whom they trust/distrust, preserve privacy and
allow the rating of any resource.
|
|
|
|
Advantages for the user
|
Reptile provides a way to subscribe to news sources (channels)
syndicated from popular websites (CNN, slashdot, etc) and other
Reptile users (your friends and other popular authors).
Reptile also provides high availability of information for the
user. All articles and channels are cached locally and Reptile
also uses other peers as cache nodes. If a website goes down
you can either fetch content from your local cache or the cache
of a remote peer.
The Reptile search infrastructure provides a mechanism which
allows you to access all this information in a very powerful
manner. For example you could find all articles on all known
peers with the word 'Linux' and sort these by date found.
Reptile also integrates the concept of reputation which we
believe will dramatically increase the power of the Internet.
Users will be able to create trust relationships with fellow
peers and measure the quality of resources in a distributed
environment. Reptile integrates the Sierra Reputation Framework
which provides a simple and powerful mechanism for measuring the
quality of information.
Reptile also has a flexible network plugin infrastructure that
will allow us to operate on multiple P2P networks including
JXTA, Freenet, GNUtella, etc.
Reptile also provides an easy and flexible publication system
for authors. Users can publish their own articles and exchange
these with other Reptile users directly or have them
automatically uploaded to a rendezvous peer in order to be
further syndicated to other news services.
|
|
Breaking through censorship and disinformation
|
One of the main reasons why Reptile was created was to provide a
communications 'network' where censorship, disinformation, and
bias of the press are kept to a minimum.
With Reptile, one can subscribe to multiple news sources
including articles from large organizations such as AOL/Time
Warner, CNN, etc. One can also subscribe to the author of a
weblog living in some 3rd world country whose opinions just
happen to be very well informed.
This is why Reptile is a P2P system. Anyone can be both a
consumer and a producer within the network.
The main reason this works is that we use reputation (still
under development) to rank the quality of articles. If someone
like CNN produces an excellent and unbiased article, its
reputation will rise. If they produce a biased and unfair
article, its reputation will decrease.
|
|
Reptile is seeking developers!
|
The Reptile project is seeking experienced developers. If you
like Open Source, understand Java and XML, and are excited about
distributed (P2P) systems, please consider helping us out.
Reptile is still under heavy development. We encourage others
to get involved and give us feedback.
|
|
Details
|
- Open Source
-
Reptile is dual licensed (GPL/BSD) under the
OpenPrivacy
Licensing Terms to provide the greatest degree of
usefulness and flexibility in its use.
- Java and XML driven
-
Reptile is a 100% Java and XML based environment. Builds
are driven by
Ant,
Tomcat hosts our servlet
environment, and Xerces and
Xalan provide the XML infrastructure.
- Syndicated content
-
Reptile is backed by a syndicated content engine which
enables it to continually check for updated subscriptions
and publish content back into the system. As the
communications are abstracted, Reptile can support any P2P
network including Freenet, JXTA, Jabber, GNUtella, etc.
Adding a new network is as easy as writing a plugin.
- Personalization and Reputation Management
-
Channels, articles and indeed all objects within the Reptile framework
can be enhanced with reputation as provided by
Sierra.
Reputations provide a facility to enable feedback for the
creation, delivery and presentation aspects of each object,
as well as enabling threshold alerts and other advanced
features.
- Channel Creation (Anyone can publish)
-
Reptile users can publish their own RSS channels. Further,
as part of channel subscription and article selection, the
user may choose to publish all or part of their filtered
feeds, creating a new 'virtual RSS channel'.
- Channel Listing
-
Reptile can talk to RSS channel feeds (and OCS feed such as
xmltree, 10.am, or moreover) and list them according to
their reputation.
|
|
|