[
Lists Home |
Date Index |
Thread Index
]
Dare Obasanjo wrote:
> Anyone who thought about the claims that XML would revolutionize search
engines for more than 10 minutes would have realized how unjustified those
claims were. The MetaCrap[0] essay is an excellent summary of the exactly
why such thinking was folly. If there is anything to learn from that
undelivered promise it isn't your points (1) and (2) below but that
technological solutions that depend significantly on [excessive] human
intervention to create automated processes are doomed to failure.
>
> One end of the spectrum is the package tracking system used by UPS[1]
which although significantly dependent on human intervention to work
(packages must be scanned at every stop) do not alter the shipping process
significantly since package handlers merely have to scan the boxes they are
moving. The other end of the spectrum are things like RDF and the "semantic
search engine markup" which require significantly altering application
processes[2] while needing agreement from dozens of distinct entities even
after making these alterations.
No doubt that is true, but imagine if a single company, say GE or GM were to
declare:
Any vendor who wishes to sell *any* parts to [GE] must declare its intention
by submitting a URI to the following location ... The URI will be
dereferenced with the following header:
Accept: application/rdf+xml; application/owl+xml
the document returned must describe the intended part(s) the vendor wished
to sell. The bid will get faster/better attention if it uses terms from the
following vocabulary/ontology ...
Now say that is the start of a Google for the SW. I am not saying that this
actually _will_ work, just that it _might_ work, and there are a few [large]
companies willing to invest substantial resources into looking at this and
developing it further.
Jonathan
|
