konspire logo the power of broadcast
[a revolution in mass-scale content distribution]
main
download
[features]
channels
start a channel
recommendations
k2b_measure
hostKatcher
[docs]
the revolution
how k2b works
results
faq
k2b is like...
other systems
BitTorrent?
[community]
foundation
forums
wiki
skins
[dev]
bugs
developers
project page
coding credits
[donate]

hosted by:
SourceForge Logo
visit:
infoAnarchy Logo

konspire2b compared to other systems

Gnutella, et al.

Gnutella, and a large group of similar systems (for example, KaZaA), broadcast search queries through the network of nodes. Nodes with files matching the search query route responses back to the searching node. In other words, data sinks (downloaders) search for data sources (nodes with files) by sending out queries. If only one node has a file that many other nodes want, that node is swamped by the demand. Of course, over time, a popular file will spread to the nodes that manage to download it, a form of ad hoc load balancing does take place.

konspire2b can be thought of as "Gnutella run backwards": data sources actively seek out sinks for their files. Instead of broadcasting search requests, sources broadcast content announcements, and nodes that want the content route back responses. Since the sources drive the distribution (and there is no way for sinks to actively request a file), they can control it in a simple way that ensures a balanced load.

the trade-off with konspire2b is that sinks can only receive what the sources are sending out. This makes konspire2b good for delivering fresh, high-demand content on the day that it comes out, but bad for making an archive of older content available on demand. konspire2b is by no means a replacement for Gnutella-like systems. In fact, the two systems are perfect complements for each other.

Freenet

Freenet is a content publication and retrieval system with a focus on anonymity for both publishers and readers. Documents are "inserted" into Freenet by pushing them out into the network. Readers send requests into the network for documents, and the documents are routed back to them. Each node along the route caches a copy of the document to speed up future document fetches.

though publisher anonymity is a side-product of konspire2b's design, it is by no means the focus. konspire2b instead focuses on extremely efficient content distribution.

the push-publishing in Freenet can be seen as similar to content broadcasting in konspire2b. However, in Freenet, new documents are pushed out to arbitrary nodes, whereas in konspire2b, documents are only sent to the nodes that want them. Also, Freenet allows documents to be retrieved on demand. Content distribution in konspire2b is entirely sender-driven.

in other "data store" systems, such as Mnet, documents are pushed into the network by their publishers. However all of these systems focus on on-demand content retrieval after publication. Though many of these system use caching to achieve load balancing, demand-distribution can be relatively inefficient.

konspire2b was not designed as a replacement for distributed data stores, which are designed to retain popular pieces of content over an extended period of time. konspire2b was designed specifically for "zero day" distribution of high-demand files that will be in low demand shortly after their release.

BitTorrent

since many people see BitTorrent as konspire2b's closest competitor, we have created a separate page that compares the two systems..