This new doc explains the PakNet information mentioned in the abstract in more detail. It is hosted on Token_Word with the title "The PakNet :: Revisited", and notes some details of implementation that were not in place when the abstract was written.
Some documentation is available at http://sourceforge.net/docman/display_doc.php?docid=34057&group_id=172078 .
Also, the new site is up: http://paknet.twilightparadox.com .
We have the XML schema for PXMF done (though we're still waiting on the DTD), and our CGI script has backend code that will post and retrieve documents. We have a neat logo (as you can see on the site) and a UI mockup (as you can also see on the site), and a working version of the UI (for development purposes -- we're using it as a sort of IDE for PakNet internal code) is in progress (should be done pretty soon). We will probably post it on the site, even though it will be mostly a demo and will not in fact connect to an actual PakNet server. It may at least be of use to those of you who want to try out the PakNet development interface to see what it's like, or maybe to actually develop with it. It will come with full documentation on how to use it, and will support Python, Tcl, and Lua code, as well as text and possibly images, storing all data in a PXMF file.... read more
Hey, welcome to the PakNet project on SourceForge. This is a new project, and a rather strange one, so I'll do my best to explain in short order what this is -- I will upload a more detailed description soon though, and you should read that too if you get a chance.
PakNet is an attempt at a Xanalogical structure for the web. In other words, it is a way of storing and manipulating hyperlinked documents, like the world wide web, except closer to the Xanadu project ( http://xanadu.com ) in its goals and structure. More specificially, it supports links that go both back and forward, transclusions (links that embed sections into the original document without copying), versioning, the immutability of versions, links to sections of versions, and the idea that link code should be outside of and separate from the document itself, instead of embedded in the middle.... read more