Dear John, (01)
See below, (02)
Matthew (03)
> -----Original Message-----
> From: ontac-forum-bounces@xxxxxxxxxxxxxx
> [mailto:ontac-forum-bounces@xxxxxxxxxxxxxx]On Behalf Of John F. Sowa
> Sent: 18 November 2005 21:32
> To: ONTAC-WG General Discussion
> Subject: Re: [ontac-forum] Some thoughts on hub ontology and merging
> sources
>
>
> Dear Matthew,
>
> If everybody agreed to rewrite all their programs from
> scratch in some new methodology, it might be possible
> to enforce a totally new framework. But even then, I
> have serious doubts that the framework would remain
> stable for any length of time (i.e., anything longer
> than the time it takes to start a new project). (04)
MW: Every so often the pain of maintaining differences overcomes
the desire for independence, and I'm not saying it should be a rule,
but that it should be supported and encouraged.
>
> JS>> Merging two small ontologies is much, much easier. And
> >> more importantly, if you are only looking at a specific
> >> task, it is very likely that the subsets appropriate to
> >> the task will have similar perspectives.
>
> > MW: This sounds just fine if all you really have is two
> > programs with one interface. But when you have hundreds to
> > thousands of programs and thousands to perhaps millions of
> > interfaces, you rapidly get an unmaintainable mess at an
> > engineering level when you consider the change that is
> > going on.
>
> Yes, indeed. That has been true for the past 50 years, and
> the only solution anybody has suggested is "Make me the Tsar,
> and I'll whip anybody who steps out of line." (05)
MW: I agree that having a Tsar gives a much better chance of
success. But take what I have managed, which is to produce a 4D
ontology in an open standards environment (ISO). No Tsars in sight
there, in fact you are positively encouraged to throw in your
2 cents worth. As I said above when the pain is sufficient (and
it certainly was for that effort) then there is a willingness
to work together.
>
> But even a Tsar cannot change three things: (1) the trillions
> of dollars worth of legacy systems, (2) human nature, and
> (3) the universe. (06)
MW: I quite agree, but the death of legacy systems comes when the
cumulative addition of functionality makes the system unstable, and
a rewrite is easier than continuing maintenance. (07)
MW: One of the models I use in Shell is one for new technology
(broad sense) it is a funnel, with lots of competing ideas/products
coming in at the top, and after the technology has matured, just a few
products remain.
>
> > MW: As a result you see in practice small systems being
> > merged into larger systems with a (more) common view of the
> > world. From what I have seen, this follows the need to gain
> > a better understanding of the world to achieve greater
> > capability.
>
> I agree that we more often see small systems merged with
> larger ones than the opposite. But so far, there is no evidence
> of real progress in people achieving a better understanding of
> the ontology. As evidence, I would suggest the comments that
> Barry Smith and I made a few days ago about a proposed hub
> ontology that was significantly inferior to what Aristotle
> preached over two millennia ago. (08)
MW: Well I would not accept that because things are sometimes done
badly, that they cannot (or should not) be done. I think there is
progress, but that it would be faster if we paid proper attention
to philosophy, and this is still done only infrequently, so progress
is more painful than it could be.
>
> > MW: However, I think the integration and merging activity
> > also has its place, and is an equally natural phenomenon.
>
> I'm not against merging and integration. (09)
MW: Good. My main point here is to argue for not excluding this,
so this is good to hear. (010)
> As you know, we've
> all been promoting the idea for a long time. But there has
> been no evidence that a frontal attack is going to work
> without a Tsar. (011)
MW: I will mention ISO 15926 again as a counter-example. (012)
> And even with Doug Lenat serving as the
> Tsar at Cyc for over 20 years, we still don't see any
> significant progress toward finding the Holy Grail. (013)
MW: I think there is a tendency to underestimate the length
of the journey to anything that might be described as a Holy
Grail, but I still think we need to work on taking the next step,
rather than throw the towel in.
>
> That is why I strongly agree with Rick Murphy that a focus
> on information flow is important, along with the related
> principles that the cited:
>
> RM> The more interesting of these principles include tolerance,
> > decentralization, test of independent invention, principle
> > of least power, free extension, language mixing, and partial
> > understanding. (014)
MW: And I agree Information Flow is an important part of the mix.
>
> Merging and integration are required to achieve tight coupling,
> but not every system requires tight coupling. Loose coupling is
> much easier to build and maintain than tight coupling. (015)
MW: I agree loose coupling is easier to build very often, but I
disagree that it is easier to maintain. Maintenance is about
engineering and order, and simply applying loose coupling is
anarchic and in the long term expensive (however, frequently
the best we can do). (016)
MW: Well I don't think there is much here we really disagree
about. We both accept that a multiple ontology approach is
necessary and that merging ontologies is also desirable. What
we seem to disagree on is the practicality of merging ontologies
and the costs of maintaining interfaces between ontologies. I
guess time will tell on those.
>
> John
>
> _________________________________________________________________
> Message Archives: http://colab.cim3.net/forum/ontac-forum/
> To Post: mailto:ontac-forum@xxxxxxxxxxxxxx
> Subscribe/Unsubscribe/Config:
> http://colab.cim3.net/mailman/listinfo/ontac-forum/
> Shared Files: http://colab.cim3.net/file/work/SICoP/ontac/
> Community Wiki:
> http://colab.cim3.net/cgi-bin/wiki.pl?SICoP/OntologyTaxonomyCo
ordinatingWG (017)
_________________________________________________________________
Message Archives: http://colab.cim3.net/forum/ontac-forum/
To Post: mailto:ontac-forum@xxxxxxxxxxxxxx
Subscribe/Unsubscribe/Config:
http://colab.cim3.net/mailman/listinfo/ontac-forum/
Shared Files: http://colab.cim3.net/file/work/SICoP/ontac/
Community Wiki:
http://colab.cim3.net/cgi-bin/wiki.pl?SICoP/OntologyTaxonomyCoordinatingWG (018)
|