kanguros

At the current state, this is more a project idea than an actual project. The name is derived from the Estonian word kangur (meaning weaver in English). The name is meant to be my small homage to this country and all Estonians. I enjoyed my time living there very much (ok, I am still there, but I may leave in 2008). It is supposed to provide a totally new way to store and access data. It will be a bit similar to the idea of the semantic net. It will extend the classical way of storing data in hierarchies to store data in random graphs.

Motivation

  • Did you ever have to store music on your computer? And got annoyed that you rather would like to put it in multiple folders? One for the artist, one for the album, one for the composer, one for the interpret?
  • Or did you ever had to store your Christmas and birthday cards somewhere? Shall you store it in a folder with the person's name, the event's name (birthday, easter, Christmas, ...)?
  • The same problems occur for all other media (pictures and video) and if you start thinking also for all other documents.
  • Also not being able to store emails in multiple folders (but only one version) is pretty annoying. It would be also nice to be able to store the attachment in multiple folders without really detaching it.
  • In general it would be nice to store documents not in some folder with an is-in relation, but to be able to use any kind of relation (has category, has priority, ...).

Some of these have been addressed for special document classes.

  • googlemail and thunderbird allow tagging of your mails.
  • zotero allows tagging and categorizing of your documents, but copies these in its own database.
  • picasa allows tagging of your pictures.
  • firefox version 3 allows to tag your bookmarks.

But why can't we just mix this and use it for all our documents? Hence, we need a system which provides a possibility to store documents in a less hierarchical way.

Solution idea

A solution is to store your data as multigraphs, similar to to the storage of data in the semantic web. If you are interested, how to do this exactly, contact me. Solution on private page.