The ins and outs of the ontology engine

We make a query on our browsers, maybe a sophisticated one with a search widget. That query becomes an xml file. So if all the world of G were linked in a large graph, their long distance links would be XML files sent to URL entry points at other locations.

So, my browser sends my xml graph to sgtn://www.wiki.graph.entry. sgtn stands for semantic graph traversal network. In our grammar we are really doing a :
convolve(MySearchGraph,wiki.graph.entry)

The whole URL, RDF thing is just the mechanics of a long distance extension cord, instead of a short distance link. It is sill not an input or output. The real entry points are blobs of text. My search editor generates keyword G, with help from the search macros, but it still started with a sting of text. Ultimately all searches point to a bob of text in html format, likely.

I think we have to model the is an outs as magical. blob of text suddenly appears linked up to something! Using clicks thrus, the new blob of text eventually becomes linked in a maximally encoded ontology graph, via smaller graphs, that just appears, convolving along an atracting click thrus. In between are autonomous mid size graphs, convolving along and arranging ontologies by click thrus, and other finding nested key words in random text. The communal editor.

My current development.
I have sqlite3, command line. I have note pad. I have the M4 macro utility. That lone is enough to make very simple, elegant language to control megabytes of text. Sqlite3 is a revoutionary tool, seriously.

So I am working on this database model in which there are three variables, all table stores, all G nested graphs with terminations (not URL jumps). The three variables are table sore Gout, G1 and G2; the grammar exactly. I am testing three groups of sql statements, one for set operations, one for sequential, an one for insert. Finally there is one for insert into from select, for very fast set operations that set Gout without re-enterin the execution loop.

All in all, about 15 sql statements. My m4 macro processor can generate enough data in G format to test without the engine, Then I will set these SQL statements up as an ontology graph, put local table store the engine can initilize the SQL statements. Cuts down on lines of c code.