SUO: Unanswered Questions About SUMO Set Theory
John-
For some reason this E-mail (and a couple others) had never been sent, sorry for the 2-week delay.
-FF
At 00:38 2002-05-27 -0700, John F. Sowa wrote:
> ...
> FF> More platitudes. Got some standards wording to propose?
>
> Yes: "The SUO standard shall consist of the IFF framework
> with the SUMO and OpenCyc contributions included as content
> modules within the framework. Both the SUMO and the OpenCyc
> ontologies shall be broken into a collection of the modules
> from which each was constructed, and those modules shall be
> organized in a lattice that shows how each of them is related
> to each of the others as generalizations, specializations, and
> alternatives. The operators defined within the IFF framework
> shall be used to derive the mappings and relationships between
> the modules."
John-
This isn't standards wording. Just adding the gratiutous "shall" to a sentence doesn't magically turn the sentence into "normative wording". While the statement:
"The SUO standard shall consist of the IFF framework with SUMO and OpenCyc contributions ..."
might be a "resolution" of the SUO WG, i.e., a directive to the technical editor(s) (editing instructions from the SUO WG), these aren't standards wording because they are not "provisions" (e.g., requirements) with respect to an "implementation".
What is an "implementation" of the standard? We would know that if we had a "conformance" statement. The notions of "implementation" and "conformance" work hand-in-hand ... sometimes it's easier to approach this by starting to describe the "implementation", sometimes it's easier to approach this by starting to describe the nature of "conformance". Last (northern) fall, I suggested that we try to describe "conformance" ... we (the SUO WG) got into some lively discussion, but the results weren't clear to me, i.e., I didn't have a significantly better sense of what "conformance" was after the discussion.
So to propose standards wording, one needs to have a sense of what an "implementation" is. Aside from your wording above that would suggest combining the 2-3 documents, even if the document were combined, it wouldn't tell use what an "implementation" is, or what "conforming" means..
> Generating the standards wording is a trivial exercise. The hard
> part is to get some very thick skulls to recognize that they would
> have a disaster on their hands if they continue as they have.
Actually, generating standards wording is a very complex exercise because:
- the group needs to build consensus around the standards wording
- which depends upon the group to have, in general, a common understanding of "what an implementation is" (or, alternatively, "what conformance is")
- and it requires understanding and accommodating, as best possible/practicable, the variety of stakeholder interests
I don't see these kind of concerns addressed in your proposed standards wording ... all I see are editing directives, which have little relationship to implementation issues/concerns.
-FF
_______________________________________________________________________
Frank Farance, Farance Inc. T: +1 212 486 4700 F: +1 212 759 1605
mailto:frank@farance.com http://farance.com
Standards, products, services for the Global Information Infrastructure