Another round of subject identity crisis c.2011

Hi folks,

In familiarizing myself with the TMQL stuff, I've also had to finally break down and look seriously at CTM (after managing to successfully ignore it for years).  While I have a clear understanding of the difference in most cases between subject indicators and subject locators (seemingly "old terminology" though), I must say I was slightly confused to see that the general consensus is that PSI references  seem to only be used as subject indic^H^H^H^Hdentifiers and never as subject locators.

My question:  shouldn't this distinction really be dependent on the nature of the resource in question and not governed by a hard rule?

While it's quite clear to use a Wikipedia page as a subject indicator for the thing being discussed and the same page as a subject locator if you want to discuss the bytes retrieved from that location, I have to say I'm slightly puzzled by the use of published topic references as subject identifiers rather than as subject locators where it could be possible.

Let's say I have a topic map as follows (testing my new CTM chops):

	mytype - "MyType";
	isa http://www.topicmaps.org/xtm/1.0/core.xtm#class;
	description: "This is my world view of MyType as I understand it as of 8 Mar 2011".

And I publish it at the following location:

	http://psi.myco.com/v1/mytype

along with ensuring that it is served as a well-formed CTM document with the proper mime-types.

Clearly, I've intended to define a topic type as a topic and I've published it in a way that should make it possible for direct consumption by other Topic Maps software.  In my view, I've done everything possible to promote my lowly item identifier 'mytype' to a "proper" PSI.  Yes or no?

If "subject locators are used when you want to refer to [a] resource that is the subject"[1] then if I, in a separate map, do this:

	mytype = http://psi.myco.com/v1/mytype .

What have I done?

In my view, what I've done is pulled in the type definition defined externally into the local map so that the semantics intended by the published topic instance that declares the type 'mytype' don't have to be repeated while at the same time explicitly indicating that the ontology of the current map is tied to the external resource defined by the subject locator.

Of course, I *could* do this instead (as per [1]):

	mytype http://psi.myco.com/v1/mytype .

as naturally, since I've defined it quite precisely at that location, it is, in fact, also functioning as a subject indicator as if I had defined it using any "normal" human-consumable resource published at that location.

Why can I not leverage the semantics of topic maps in this case?  What are the pitfalls of using a subject locator IF:

1) I've made every effort to define the resource in a machine-friendly format processable by topic map implementations, and

2) I've made every effort to ensure that said resource is served in a way that would provide all possible meta-data to indicate that this resource WAS a valid topic maps construct rather than any arbitrary, untyped resource which could be retrieved from that location.

If you can't do this to ensure you are sharing the ontology definition itself rather than references to those ontology definitions, then I think we're losing some potential value here.  As I under stand it, any compliant topic maps system sees subject identifiers as simply opaque strings which are applied in the particular context of establishing identity, but that's it (this is certainly what I do).  It would be beyond the scope of the definition of a subject identifier for the topic map software to make any attempt to process it.  Naturally, applications *using* the topic map software are a whole different ballgame.

Is explicit merging really the only way to effectively accomplish re-use of topic definitions, be they types or otherwise?

Should we have a way to provide semantically valid references to topic maps constructs external to the current map?  I would certainly think so, but while I thought I understood a way in which this might be accomplished (admittedly by pushing the boundaries a bit as I've discussed before[2]), in trying to apply this to illustrate how I would solve the CTM use cases using my own system and notational conventions, I'm not not sure.

Any ideas, thoughts or discussion of the above would be appreciated.

Cheers,

ast

[1] http://www.topicmapslab.de/articles/clarification_regarding_identifiers
[2] http://groups.google.com/group/topicmaps/browse_thread/thread/e99f7f43e82720d1

BTW, has this always been a google group?  I think this is the first time I've ever seen an archived post show up there.
--
Andrew S. Townley <[email protected]>
http://atownley.org