[sc34wg3] New SAM PSIs

[Lars Marius Garshol]

* Mary Nishikawa
| 
| [subtype loops]
|
| Can you give an example of this? 

I don't know of any, admittedly.

| I do not think that this works. You can create a loop, but does it
| have any bearing to reality?

Well, what you are saying with the loop is basically this: "these
classes have different definitions, but they all have the same set of
instances". There's nothing unreasonable about that.
 
| Please convice me with a real example.
| It should work in all cases of super-sub and if it doesn't then we
| need something else to describe this looping.

I think the burden of proof should be the other way around, actually.
What we are discussing is whether to add a constraint to the SAM
(disallow subtype loops). If we do this we are putting a burden on
implementations (they must check for and report such loops) and we
will also be preventing applications from doing something that for all
we know may be useful. (In fact, we already know that the knowledge
representation community does this.)
 
| Of course if A=country and B=city, this can never work.

Sure. In most cases it's nonsensical.

* Lars Marius Garshol
|
| DAML+OIL defines the equality between classes that way.
 
* Mary Nishikawa
|
| Please give me a reference.

<URL: http://lists.w3.org/Archives/Public/www-rdf-logic/2001Sep/0001.html >
<URL: http://ilrt.org/discovery/chatlogs/rdfig/2001-11-16.html#T20-37-01 >
<URL: http://www.w3.org/TR/owl-ref/#equivalentClass-def > (note at bottom)

| Yes, if a subclass = superclass and a superclass = subclass then they
| are the same class.
| and the distinctions of being super or sub disappears.

Not necessarily. If you read

  <URL: http://www.isotopicmaps.org/sam/sam-model/#subtype >

carefully you'll see that if we have A and B are in a subtype loop we
are allowed to infer that every instance of A must also be an instance
of B, and vice versa. (In other words: they must have identical sets
of instances.) It does *not* allow us to infer anything else.

The logical conclusion is that whoever asserted that loop wanted to
say that the instances were the same, but that the classes were
different. 

-- 
Lars Marius Garshol, Ontopian         <URL: http://www.ontopia.net >
GSM: +47 98 21 55 50                  <URL: http://www.garshol.priv.no >