[higgins-dev] The case for "node" (RE: Draft ITU-T Report on definition

["Drummond Reed" <drummond.reed@xxxxxxxxxxxx>]

The message below (from a thread this morning on the IDGang list on the term
"identity") summarizes the core problem with using the term "entity" to talk
about the "digital representation of an entity":

At 08:54 PM 3/1/2008, Dave Kearns wrote:
>"Digital Identity" or "Persona" are the two best candidates for the digital
>representation of an entity. Why anyone in their right mind can think that
>the dual use of Entity is "self-evident," I don't know.

Although there may not be agreement on this, I think we would all
acknowledge that using the same term to refer to both the representation and
to the thing-being-represented is less than ideal. What's worse, of the two
potential meanings, the term "entity" has now been defined by two bodies --
ITU and IDGang -- to explicitly mean the thing-being-represented, whereas
the term the Higgins community needs to standarize on is the representation.

If that disqualifies the top choice in the last poll -- "entity" -- and
we've already disqualified "digital subject" due to the legal/political
issues -- then that leaves us with the second choice in the last poll:
"node".

Let me briefly summarize the case for "node":

1) It doesn't suffer from the legal/political connotations that disqualified
"digital subject".

2) It avoids the semantic confusion of either "entity" or "identity".

3) It is much shorter and more concise than "digital identity".

4) It corresponds directly to the same term in the RDF graph model that
underlies the Higgins data model.

5) It emphasises the power of the Higgins global graph model (and the
related concepts of the "social graph" and "giant global graph"), in
particular for the relationships between real-world entities to be
represented by relations/correlations between nodes in different Higgins
contexts.

Lastly, the first sign I look for regarding the viability of a new term is
how easily it works in everyday usage. So far my personal experience is that
"node" is working very well. Several of us were able to use it in a
first-time meeting yesterday with a potential large Higgins adopter with no
definition or explanation at all -- it just flowed naturally from the basic
premises of the Higgins data model, which is the foundation upon which
everything else is built.

My conclusion is this: while I fully appreciate what Raj and others have
said about the initial "geekiness" of the term "node", I believe that this
connotation will quickly disappear with Higgins adoption, very much the way
the initial "bookishness" of the term "browser" disappeared quickly with the
spread of the World Wide Web.

The result will be that Higgins will have contributed two fundamental terms
to the industry-wide vocabulary of digital identity and data sharing:
"context" and "node" -- arguably the two most fundamental terms in Higgins
architecture (along with "attribute", which is the only holdover from the
LDAP paradigm).

So I'd just like to cast a new vote for going with "node".

=Drummond