Page 1 of 1

ISO 15926-7 Template Methodology

Posted: Fri Sep 28, 2012 3:41 am
by Andreas
Hi All,

I would like confirmation on whether:
1. Users are only permitted to construct specialised versions of the templates comprising the initial set?; OR
2. Whether users are permitted to construct their own templates (that are not specialisations of the initial set) AND begin using them immediately?; OR
3. Any new non-specialised templates must go through an approval process before becoming part of the standard?

Regards
Andreas

Re: ISO 15926-7 Template Methodology

Posted: Fri Sep 28, 2012 9:16 am
by OnnoPaap
Hi Andreas,

In my opinion the short answer is option 3.

Here is the long answer.

The first question is who are you addressing? When you say "permitted" do you imply an institute overseeing the 15926 workers? Or do you imply a rule: if you want to be 15926 compliant you are not permitted to...

The procedure set up by the JORD project of handling new or revised base and core templates will be the compliancy spec.

Until that exists (and also maybe after that) we use the opinion of the MMT SIG as basis. The MMT SIG is the forum group in this link
Andreas wrote: 1. Users are only permitted to construct specialised versions of the templates comprising the initial set?; OR
The initial set is replaced by the proposed template list in combination with the prototemplates

So there is no "approved" template yet. They are all proposed (there will be a column added to the list for the status)
Andreas wrote: 2. Whether users are permitted to construct their own templates (that are not specialisations of the initial set) AND begin using them immediately?; OR
In my opinion there are two answers to that.
  1. In normal circumstances, like working with iRING, the answer is no. The templates used must be approved and you will have to use approved Object Information Models (OIMs) to be compliant and to reach interoperability. The 15926 is an upper ontology and we don't believe in mapping before inter-operation. We believe in pre-mapped constructs that will automatically bring data integration without requirement of re-mapping.
  2. But with the proper software, any template you invent will bring certain semantics when lifted using the (First Order Logic) axioms, and can automatically be compared for interoperability to the lifted templates of somebody else. The premises is that:
    • The lifted instance model is correct (this is very hard to do)
    • There is such a performance that lifting is not a problem (that must be a special kind of software)
    We don't believe this state can be reached in "normal circumstances" so for the time being (or forever) we stick to approved base and core templates and centrally offered (JORD servers) global OIMs.
    However a federation of 15926 servers, maybe working on the same (short scheduled) construction project in a secured ring, can always declare one of their servers the standard templates and OIMs, and then for that federation interoperability is reached.
Andreas wrote: 3. Any new non-specialised templates must go through an approval process before becoming part of the standard?
In my opinion, after considering above thoughts, for the time being or forever, this is the correct answer.

Re: ISO 15926-7 Template Methodology

Posted: Sat Sep 06, 2014 9:43 am
by alice
I feel there are really no such things as p7tm:Template or p7tm:BaseTemplate. They are exactly the same things as p7tm:TemplateStatement and p7tm:BaseTemlateStatement. Their introduction as separate level in classification chain really breaks Part 2 typing.

Re: ISO 15926-7 Template Methodology

Posted: Sat Sep 06, 2014 10:38 am
by OnnoPaap
Hi Alice. Welcome to the 15926.

This thread is very old. Almost 2 years.

Since then we have had a redefinition of the p7tm model and then another one.

Eventually Hans came out to this
http://15926.org/publications/templates ... /index.htm

So this is the current proposal to be the basis of the core templates that are here
http://15926.org/15926_template_specs.php