Building pleasant and stable islands in a storm-tossed sea ...

Traction Roots: A Whirlwind Tour (.ppt 6.2MB) tells the Traction story in pictures: 1) Tim Berners-Lee's web trades stable links for utmost simplicity and bottom-up scalability without central control; 2) Traction creates spaces which are pleasant and stable islands with a rich hypertext model internally: bi-direction links; comments based on ternary relations rather than hacking the representation of the referent object; faceted permission models uniformly enforced for search results, cross-references, as well as content browsing; fully journaled actions, etc. 3) Traction generates HTTP addressable views of its content to enable any item in the Traction corpus to be read and linked like the rest of the web (optionally restricted by access controls). This creates a pleasant and stable island that's easily connected to other islands of stability on the Web - as well as anything in the storm tossed sea - not a stovepiped box.

We set out to build a hypertext system that could natively link to anything and interoperate with anything on the Web, rather than limiting the domain of discourse to whatever people chose to store within a proprietary hypertext box (be it NLS, Intermedia, or Lotus Notes).

Although Traction's logical schema is a rich hypertext model internally, it uses pluggable skins to render the content as permission filtered HTML or XML views on the fly when communicating with Web browsers, RSS readers, search engines, or other agents. That's where a lot of experience and IP resides as well.

We also support interoperability through a choice of editing interfaces: Zero-footprint Ajax style editors deployed in a browser; Metaweblog or other common API's for blog posting/editing tools; a .net client / IE plug-in for drag-and-drop integration with the Windows desktop. Apollo and Silverlight will open up very interesting new possibilities without sacrificing the ability to work as part of the Web.

You'll be able to get an editing/viewing interface that's as rich as you want, combined with HTML / XML views of the same corpus to make it linkable at a fine grain using standard W3C protocols. Because we mediate the edits, the rich hypertext model retains its internal integrity and presents a stable view to any agent: rich client, Web browser, syndication reader, Web search engine, etc.

A more complete version of the story is in Blog104: Use of Weblogs for Competitive Intelligence | First International Business, Technology CI Conference, Tokyo Oct 2005

See also Public390: 20 June 2005 | Supernova | Why Can't a Business Work More Like the Web?
TeamPage 4.0 | ハイパーテキストを活かしたビジネス向けプラットフォーム
Public390: 20 June 2005 | Supernova | Why Can't a Business Work More Like the Web?
Public219: 16 March 2005 | Blogs, Wikis, and Beyond: New Alternatives for Collaboration and Communication
Blog106: The Evolution of Personal Knowledge Management