The Many Sides of the Internet: Arpanet, Darknet, and Others



The Internet is a familiar concept to numerous people. What if the Internet is different than what people thought? That is, what if there is more to the "Internet" than what most people see and know? To the surprise of many, this is true. The Internet has multiple parts or sides. The side many people use and know about is called the ARPANET or Surface Internet. However, there is more.

General Concepts
The ARPANET, Surface Internet, Clearnet, or Surface Web is the most commonly used and known World Wide Web. Most search engines can crawl and index this portion of the World Wide Web.

The AMPRNet (AMateur Packet Radio Network) is a private subnet on This subnet is similar to the concept of "amateur radio".

The Dark Web is the portion of the World Wide Web that is unknown to many people. Most search engines cannot crawl or index this portion of the World Wide Web.

All portions of the World Wide Web that is not indexed or crawled by search engines is called the Deep Web, Hidden Web, Invisible Web, or Deepnet. The Deep Web overlaps all areas of the World Wide Web. The opposite (or reciprocal) of the Deep Web is called the Lightnet, Indexable Web, Clear Web, or Indexed Web; search engines can crawl this web. Most of the Dark Web is also the Deep Web, and most of the Surface Web is also the Clear Web.

Dark Webs
The Darknet (Tor network, Onion Web, or Tor Web) is a network in the Dark Web that offers security and privacy. To access the Darknet, the Tor-Browser is required. Alternately, TorChat ( can be used for Instant Messaging (IM) over the Darknet. Orbot ( can be used to tunnel Android network connections through the Darknet.

The Darknet is designed a little differently than the Surface Internet. Many websites on the Darknet use domain names that are 16-character hashes (random characters) that make up an 80-bit number in base32 rather than purposely designed domain names (like "" or ""). Also, many website domain names (or addresses) end in ".onion" instead of ".com" or ".net". JavaScript is typically disabled in the Tor browser because Darknet websites are not supposed to use JavaScript. This helps with security and privacy. The Darknet has its own set of search engines since Google and Bing are designed for the Surface Internet. On the Darknet, the search engines only search the sites that want to be searched. Three popular search engines for the Darknet are Torch (http://xmh57jrzrnw6insl.onion/), DuckDuckGo (http://3g2upl4pq6kufc4m.onion/), and Tor Search (http://hss3uro2hsxfogfq.onion/) which is also called "Not Evil". More can be learned about Tor at

Tor2Web ( is a proxy on the Surface Web that allows regular web-browsers to access the Darknet. Type the address of the desired Tor website in the browser's URL-bar, but replace ".onion" with "". The Darknet can connect to the Surface Web because the last onion router (exit node) connects to the Surface Web server. To specify an exit node to use when connecting to a Clearnet website, use the address format "clearnet_hostname + .exitnode + .exit". The ".exit" pTLD is used on the Darknet for onion routers that act as exit nodes.

NOTE: Despite the Darknet having its own set of search engines, it is still considered a Deepnet because traditional (or public) search engines cannot access Darknet websites.

WASTE is a now a defunct portion of the Dark Web that was similar to the Tor Web.

NOTE: ICANN ( manages the global Domain Name System (DNS), the DNS Root Zone, and generic top-level domains (gTLDs). The Internet Assigned Numbers Authority (IANA - manages all other top-level domains except pseudo-top-level domains (pTLDs) such as ".onion". Pseudo-top-level domains are managed on non-global DNS servers such as the ones on the Darknet and GNUnet.

GNUnet ( is an open-source decentralized peer-to-peer (P2P) networking framework. This framework offers privacy and censorship-free file-sharing. Many communication protocols and transport methods/algorithms are supported as well as link encryption. Files that are shared on the GNUnet are ECRS (An Encoding for Censorship-Resistant Sharing) coded prior to traveling on the network. GNUnet supports two network topologies - mesh and friend-to-friend (F2F). Also, GNUnet uses Uniform resource identifiers rather than IP addresses. In addition, GNUnet uses the GNU Name System (GNS) instead of the Domain Name System (DNS). This allows users to create their own master zone. The GNS server maps the data into the DNS namespace under the ".gnu" top-level domain. The GNS lookups are completed by DHT (part of GNUnet). All of these features make privacy and censorship-resistance possible. GNUnet addresses begin with "gnunet://". Server domain names end with ".gnu" and zones can be accessed directly by using the ".zkey" pTLD. GNUnet is part of the Lightnet because search engines can use the DNS2GNS gateway to access and crawl GNUnet websites.


Freenet ( || is an open-source decentralized peer-to-peer (P2P) networking framework. Freenet offers protection from censorship, and it uses a decentralized distributed data storage. Freenet can also publish "freesites"; a freesite is a website that is only accessible through Freenet. Freenet is cross-platform and is written in Java. The "Freenet" is similar to the regular Internet. However, Freenet (the program) uses different protocols and algorithms to access the Freenet (network) than what regular browsers use to access the Internet (or "Surface Net"). Using a typical web-browser, people can access the Freenet via Fproxies which are proxy servers that connect the Surface Web to the Freenet. Some Fproxies include and .

NOTE: A large list of DarkWeb links can be found at

Another alternative to Freenet and GNUnet is the Invisible Internet Project (I2P - || This software is open-source and cross-platform and supports file transfers, websites, chatting, email, etc. (just like Freenet). Websites on the I2P network (not part of the Surface Net) are called "Eepsites" and their URLs use the ".i2p" pTLD. Instead of DNS, I2P uses "netDB". I2P tunnels data through "garlic" which is a technology similar to Tor's onion protocol/routing. Orchid ( is an I2P plugin that allows I2P software to support ".onion" addresses, thus making the I2P client act as a Tor client. However, I2P does not tunnel or route on the Tor network.

NOTE: For other I2P plugins, checkout

The Kad Network is P2P and uses the Kademlia P2P overlay protocol.

Torrents and File-Sharing
OneSwarm is a cross-platform P2P torrent client written in Java that provides privacy for file-sharing (

Tribler is another cross-platform P2P torrent client ( However, Tribler uses onion routing (also called the "onion protocol").

iMule (invisible Mule) is an open-source and anonymous P2P software which connects through I2P and Kad. iMule can be obtained from www.imule.i2p .

  • World Wide Web (Global network)
    • Surface Web (ARPANET or Clearnet)
    • Dark Web
      • Darknet
      • WASTE
      • GNUnet
      • Freenet
      • I2P
  • Clear Web (Indexed Web, Lightnet) - indexed and crawled by search engines
  • Deep Web (Hidden Web, Deepnet) - not indexed or crawled
Many other peer-to-peer (P2P), friend-to-friend (F2F), overly, anonymous, and pseudonymous networks and webs exist.

Further Reading


Last edited:

Members online

Latest posts