Universal Catalog

A universal catalog is a system embodying a few-to-many publication model aimed at disseminating metadata and location information for static content.

What that means is that if you're looking for some, possibly unspecified, static object (e.g., a novel, movie, music, software) then you'll be able to find information about that object in a universal catalog, using a combination of browsing and searching. In addition, your client might find a repository for that object. If it does, it will then attempt to retrieve it automatically, and it shouldn't matter a whit whether that repository is a web site, bittorrent peer, or Usenet over carrier pigeon.

This is superficially similar to IMDB, Amazon, library card catalogs, and AudioGalaxy?. However, to be worthy of the name, a universal catalog must also adhere to the following principles:

Universality and Openness, unlike the above services.
Unity of Content, unlike Usenet, IRC, etc.
Freedom of Publication, unlike Wiki.
No central point of control/failure, whether physical (storage) or legal (company), unlike Wiki.
Resilience against attack and sloth, unlike web novel archives.
Accessibility and Usability, unlike libraries.

To accomplish the above, a universal catalog must be massively distributed, ObjectOriented, versioned both inside and out, fault-tolerant, completely automated, searchable, easy to use, have bidirectional references, and a next-gen interface.

Universal catalogs were conceived and designed by RichardKulisz. An implementation called Objects is now currently being implemented, also by RK, both for its own sake and as a stepping stone in a comprehensive plan to remake the entire computing experience from the top down. The long-term goals for Objects itself include the destruction of Hollywood and the shrink-wrap software industries by making seamless distribution completely effortless.

Architecture

3G UI, a DirectManipulation framework (but will use a shit UI to begin with).
book, author, category, keyword, content, and search objects (will have a fixed ontology for now).
log; reverse maps; repository index; content; request stack.
log, maps, index and content distribution (FreeNet model limited to groups); log writing proposals, and arbitrations (piggy-backed over Usenet); groups management (group memberships, groups list, membership lists, export lists).

See UnversalCatalogTalk?

CategorySoftwarePolitics CategoryProject