Hacker News Reader

One issue that intrigues me about IPFS is how we are going to search for the data that is already available on the network?

Assume a new node joins the network and has no prior knowledge. How would this node search for files related to concept X? How could he find out which files are already available?

View on HN

Directory listings to other directory listings to eventually files. Just like walking old FTP sites or web sites. IPNS gives top level names for some of the points of interest.

Eventually someone may build tools like early FTP's Archie search tool or Gopher's Veronica search tool. A lot of the "Linked Data" proposals are reborn versions of Semantic Web search tools that might lead to interesting results (though arguably the Archie/Veronica model seems easier).

The search is done by searching the hash function of the object. When a new user views an object it is saved as cache on their node. As files are viewed more they are cached on more nodes which is similar to a built-in High Availability.

Reply on news.ycombinator.com