Finding Out About [R.K. Belew, Cambridge Univ. Press, 2000] has an interesting take on open source. Belew allows registered users of his site/book to vote on which open source organizations should receive part of the proceeds from the book.

Hope this idea catches on.

It is possible the tools Belew supplies would be pretty useful for bloggers -- access to the Google API is great, but it could be pretty useful to build custom indexes on blogs and/or blog references.

"The word BRICO is a shortened version of the French word bricolage, which denotes a functional but haphazard assemblage of components that solve a problem. The spirit of BRICO is the assemblage and interconnection of knowledge resources, rather than the characterization of any over-arching primitives or principles. We believe that this sort of architecture is more like the human mind than the elegant logical lattices of traditional knowledge bases. And we believe that we will learn more about human knowledge by constructing such a composite knowledge base. "

One notion is that translations of documents can be used to constrain meaning in document because a human has already made a commitment. That already suggests something of an exemplar based ontology. Generative Programming's Appendix A gives some background, but don't expect a close relationship between GP and BRICO.

SOAP hackers, SOAP tool vendors, and hardware companies better hope XTalk and Vinci does not catch on. Skills will be devalued; complexity will be reduced; electric power will go unused. The authors admit that untyped data will wander about the net:

"Note that unlike SOAP with its default encoding rules, there is no explicit typing of data in Vinci messages. There are two reasons for this approach. First, the sending of typing information in every message is an unnecessary overhead since this information is already implicit in the input and output schema definitions of the service, of which we assume the client is knowledgeable. Second, including typing information within each message complicates the evolution of both clients and servers because changes in this information must be synchronously distributed and incorporated into their implementations.

"Vinci clients and services are, in contrast, expected to access their document models declaratively instead of immediately forcing them into some rigid typed structure. By declaratively, we mean through non-positional, associative access, e.g. an XPath expression involving the tag names of interest. In effect, any tags present in a document that a server or client does not understand will be ignored. "

;)