Dear Matthew and Henson, (01)
This thread raises some important points: Nearly every large system
is unique, there are many common patterns, "local things" pop up in
unpredictable ways, and working systems are revised and extended
throughout their useful lifetimes. (IBM uses the term 'functionally
stabilized' as a euphemism for 'obsolete'.) (02)
HG
>> If the diagram is a specification for which there may be many
>> or no realizations... (03)
MW
> Distillation units are generally one-off constructions, there
> may be some common features, but things like the local climate
> and availability of cooling water, and its temperature, will
> all affect the design. But there is a design as well as the
> individual even so, just only one implementation. (04)
Note those local "things" like "climate and availability of cooling
water, and its temperature". Just consider how a common pattern will
change from the tropics to the arctic, from a desert to a swamp, and
from solid ground to a ship or a plane. (05)
Linux has been implemented on cell phones and supercomputers, but those
local things can create huge differences. (06)
HG
> For example one has an ontology (pattern) for an aircraft decomposition.
> These patterns can greatly accelerate product development...
> I haven’t noticed any comments regarding the formulation of patterns,
> or for that matter any substantive comments on the whole dialog. (07)
I agree that the essence of ontology involves the discovery of patterns
and their formulation as generalized, reusable formulas or templates. (08)
HG
> An engineer friend of mine in charge of a line of semiconductor manufacturing
> systems (very large, complex and expensive) pointed out that when they made
> a service call for a customer’s system, the first thing was to identify what
> had been added or changed since it was delivered. These are as you know
> the real world of engineering and any useful formalism has to address this
> kind of thing. (09)
I very strongly agree. (010)
MW
> Actually two plants genuinely being identical in specification is
> surprisingly rare. I can only think of 2 occasions where this has happened.
> In both cases the plants were built next to each other. A much more usual
> circumstance is that a design is evolved and developed from a previous one. (011)
That observation is just as true of computer systems. Developers are
more likely to learn from and imitate patterns they see in a successful
implementation than in somebody's abstract ontology. (012)
In the thread "What goes into a Lexicon?" in Ontolog Forum, I made
a recommendation that is just as relevant to this thread: (013)
JFS
> Recommendation: Instead of developing "proactive" standards for
> ontology, I suggest that we note the law of standards: examine
> what actually works, harmonize the best practices, and build the
> new additions as extensions to the de facto standards. (014)
The law of standards was explained in the following excerpt. (015)
John
_______________________________________________________________________ (016)
In 1991, I formulated a "Law of Standards" that is repeatedly confirmed
by every group that tries to develop a so-called "proactive" standard: (017)
Whenever a major organization develops a new system as an official
standard for X, the primary result is the widespread adoption of
some simpler system as a de facto standard for X. (018)
See http://www.jfsowa.com/computer/standard.htm (019)
Another confirmation of the law of standards: The W3C proposed OWL
as "The Web Ontology Language" and RDF triples for the data. The
results: (1) The overwhelming number of OWL ontologies published
on the WWW are limited to the subset of logic that Aristotle
defined for his syllogisms. (2) JSON has become the de facto
standard that programmers use instead of RDF. (3) The schema.org
developed by Google, Microsoft, and Yahoo! replaces RDF and OWL
with the simpler de facto standards of JSON and Aristotle. (020)
_________________________________________________________________
Msg Archives: http://ontolog.cim3.net/forum/ontology-summit/
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontology-summit/
Unsubscribe: mailto:ontology-summit-leave@xxxxxxxxxxxxxxxx
Community Files: http://ontolog.cim3.net/file/work/OntologySummit2012/
Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2012
Community Portal: http://ontolog.cim3.net/wiki/ (021)
|