Wednesday, December 7, 2011

Changing name space

The ontology folks do good work, I have bought into their triple definition, and thus assign it to them. The code name changes will reflect that, and triple no longer belongs to the machine, it is now an higher ontology definition. How is the new triple defined in machine language. First lets look at the real triple:
[subject][predicate][object]

Typeface Explosion can carry this load, by reserving the link field. In g all arguments and results are nested g graphs.

g = Subject.[Overload operation](for this sub graph h)

The bracket is where the overload syntax is needed. We have three bytes in link to make it happen. Solution: The machine should always reference the Dublin Core attribute set, so how is it expressed? in TE? I say use the dollar sign, it appropriately startles the client that he is specifying an attribute.
Subject$BuiltFromThese(Object1,object2.object3.object4,object5)

BuiltFromThese is from locally known set called the core attribute set. Every machine carries a core look up to move from attribute ID to attribute text identifier. So we have the real standard triple format. On match, a call back may want to reprocess the whole enclosed collected sub graph, it can be directed by the overload operators. On the command line, local translation is simple.

Te class definitions can utilize the schema operator and overload that.

What do we call g nested order and its operators? Triplets! That way it is clear, the intent is to support higher and more compresensive ontology linkage.

No comments: