[tmql-wg] subjectIndicatorRef and resourceRef in TMQL

Rani Pinchuk Rani.Pinchuk@spaceapplications.com
Wed, 17 Mar 2004 13:50:25 +0100

Hi Lars,

Thanks for the hint :-)

I have read the TMDM before. But that document was quite tough for me
:-) I looked at it again now, and realized that role, type and scope are
always given as topic items. Till now I assumed (wrongly) that XTM 1.0
follows the TMDM. Can I assume that XTM 1.1 follows the TMDM?

In that context (of XTM 1.1), I would like to ask if there is a mistake
in the DTD of it (Appendix B) when defining the subjectIdentity. That
definition looks like it contradict the definition in section 4.5.

I have several questions about TMDM and XTM 1.1. I am not sure which
mailing list is the correct one for those questions. So I am sorry if it
is a bit off-topic.

1. Why in XTM 1.1 the variants can be still placed one inside the other,
while in the TMDM they are flattened? Is this a backward compatibility

2. I saw that there is in TMDM type for each base names but not for the
topic. However, in the XTM 1.1 there is instaceOf for both the topic and
the baseName. Is that a backward compatibility issue? 

3. Should we understand the type(s) of a topic as the type(s) of its
base name(s)? If not how should we treat for example in the use
cases document? 



On Wed, 2004-03-17 at 09:15, Lars Marius Garshol wrote:
> * Rani Pinchuk
> | 
> | Is it true to say that the current TMQL candidates don't try, for
> | example, to find a base name with a scope which is defined by
> | resourceRef or subjectIndicatorRef tags?
> | 
> | Can someone explain why? 
> This has to do with the difference between the data model and the
> syntax. (I get the feeling you would really benefit from studying the
> TMDM and the XTM 1.1 specification.) Any <resourceRef> or
> <subjectIndicatorRef> found outside <subjectIdentity> is effectively
> the same as
>   <topicRef xlink:href="#foo"/>
>   ...further down (or higher up)...
>   <topic id="foo">
>     <subjectIdentity>
>       ... subjidref or resref here...
>     </subjectIdentity>
>   </topic>
> | Do we expect that the TMQL engine will always reify scopes,
> | instanceOf (of topics, occurrences and associations), members and
> | roleSpecs when parsing the XTM?
> Yes, that's pretty much it.
> One way to learn how this works may be to download the Omnigator and
> use the Canonical XTM export option there, which will show you the
> data model instance created for that topic map quite clearly.