View unanswered posts | View active topics It is currently Mon Jul 23, 2018 6:06 am



Reply to topic  [ 10 posts ] 
 Templates OWL file 
Author Message

Joined: Sun Jan 22, 2012 9:14 pm
Posts: 189
This topic is about problems with the templates OWL file coming out of the template list (link at the top)
at: 15926_template_specs.php


Tue Sep 24, 2013 1:01 pm
Profile

Joined: Sun Jan 22, 2012 9:14 pm
Posts: 189
Andrew Prosser wrote:
Onno,

My understanding of template definitions from Part 8…

Subclass of : BaseTemplateStatement
Type : Template

Why do the latest definitions not follow these rules?? It means people have to keep fiddling their tools

Thanks
Andy


Tue Sep 24, 2013 1:02 pm
Profile

Joined: Mon Feb 27, 2012 11:01 pm
Posts: 282
Location: Moscow, Russia
Namespace dm: for Part 2 entity types should be uniformly defined for all PCA RDL reference data items. If currently proposed entities are submitted to PCA - there are no other options then http://rds.posccaesar.org/2008/02/OWL/ISO-15926-2_2003#

The only publicly available OWL models of Part 2 ontology available from https://www.posccaesar.org/wiki/ISO15926inOWL are using this namespace. To allow any reasoning or verification - we need the same data model entities, not any other.


Tue Sep 24, 2013 1:18 pm
Profile WWW

Joined: Sun Jan 22, 2012 10:02 pm
Posts: 283
Nine months ago I filed my suggestions to improve on Part 8 regarding the template model: http://www.15926.org/publications/templates/template-model/index.htm.

Parts 7 and 8 are TS's - Technical Specifications - and the idea of that is that the community starts implementing and discussing, and after three years start the official ISO Ballots with a DIS (Draft International Standard) version.
In the Foreword of Part 8 you can find:
Quote:
An ISO/PAS or ISO/TS is reviewed every three years with a view to deciding whether it can be transformed
into an International Standard.

If any suggestion is totally ignored, no such DIS will evolve.

The Part 8 template model has its flaws, and we need to discuss this in the MMT.


Tue Sep 24, 2013 1:41 pm
Profile

Joined: Mon Feb 27, 2012 11:01 pm
Posts: 282
Location: Moscow, Russia
Hans,

ISO processes here come into obvious contradiction with Semantic Web practices. I don't know really - whether there is any other ISO standard which is making use (and standardising!) URIs and ontologies as we are?

It may be an easy task to change ISO TS in 3-5 year timeframe.

But on the SW once you've published any URI - it is there for eternity. In the world of ontologies rules are similar. You can not completely rebuild class structure for a public ontology without offering a mapping to a previous version.

So we've to weight our options. What are the reasons behind the changes?

Can not your proposals concerning template model be rewritten preserving initial semantics for Template and Template Statement classes at least?


Tue Sep 24, 2013 1:57 pm
Profile WWW

Joined: Sun Jan 22, 2012 9:14 pm
Posts: 189
In my opinion, using the address http://rds.posccaesar.org/2008/02/OWL/ISO-15926-2_2003# is not logical.
It is a URL. It can easily change.

For example, the posted file is not dereferenceable and cannot be used in SPARQL queries either.
Once somebody decides to load the file into a triple store the endpoint address is the new location so what will you do?

I propose to use the URI: http://standards.iso.org/iso/15926/dm/ and to put a redirect header at that location to wherever we choose.


Tue Sep 24, 2013 2:26 pm
Profile

Joined: Sun Jan 22, 2012 10:02 pm
Posts: 283
That URI leads to:

Attachment:
iso-portal.jpg
iso-portal.jpg [ 56.33 KiB | Viewed 2156 times ]


Is that what you want?


Tue Sep 24, 2013 3:00 pm
Profile

Joined: Sun Jan 22, 2012 9:14 pm
Posts: 189
I have no control over that location, nor do I know who does.

At that location we need folders dm, tpl, tm, rdl etc who each have a redirect to a Sparql endpoint.

We need to set up the endpoints, test the redirect, and change the redirect when required.

Technically, what we need is a 301 redirect which can be made with Apache rewrite_module or IIS redirect setting. It is simple, really, and not very heavy to the iso server.

So we need some level of control.


However, none of this has to do with the change management process we need on updating our TS (of part 8 in this case) which will notify users of the standard who can respond. For which this thread was started.


Tue Sep 24, 2013 3:15 pm
Profile

Joined: Sun Jan 22, 2012 10:02 pm
Posts: 283
Whois.com tells me that ISO is the Registrant.

W3C tells us that ontologies must be updated in versions, so not one bit at a time.
That means that, whoever controls those versions, must have full access rights.

I think that T25/WG3 must supervise that process and organize the party doing the work and organize those access rights.

I don't think that ISO will be capable, nor willing, to provide a service with 99.999 % uptime.

That may not be necessary, because you can download the entire contents in your local server, and link those to you local ontology by means of owl:imports (if your triple store allows for that).


Tue Sep 24, 2013 3:35 pm
Profile

Joined: Mon Feb 27, 2012 11:01 pm
Posts: 282
Location: Moscow, Russia
Onno, there are several separate topics.

Quote:
In my opinion, using the address http://rds.posccaesar.org/2008/02/OWL/ISO-15926-2_2003# is not logical.
It is a URL. It can easily change.


It can change as easily as any other technical system in control of PCA. If we're working together to do better PCA RDL - this can not be an argument in our discussion.

Quote:
For example, the posted file is not dereferenceable and cannot be used in SPARQL queries either.


You're mistaken here. These URIs (not the file!) are dereferenceable. Put them in a browser. The file opens and it is the legitimate way to dereference hash URIs, as described by our W3C gurus.

Quote:
Once somebody decides to load the file into a triple store the endpoint address is the new location so what will you do?


Nothing will change. An address of an endpoint has nothing to do with address of namespace destination. You'll be able to query the same relationships and get the same triples from endpoint, and URIs will remain dereferenceable to the same location. The owner of an endpoint will be wrong to change content of this ontology without changing the file at PCA site, of course.

Quote:
I propose to use the URI: http://standards.iso.org/iso/15926/dm/ and to put a redirect header at that location to wherever we choose.


That will change everything and render my argumentation void. Then we can reconsider this choice, wait till namespace is changed in the whole of PCA RDL(!) and adopt it ourselves.

But we can not wait till that moment. Again, the moment first of Hans' templates is approved and published as PCA RD item - everything will be completely screwed up if dm: namespace is inconsistent. We can wait till this moment and use whatever namespace you like in sandbox (I'd like a standard one as a non-standard is ruining my verification attempts, but I can change it manually at least).


Tue Sep 24, 2013 3:39 pm
Profile WWW
Display posts from previous:  Sort by  
Reply to topic   [ 10 posts ] 

Who is online

Users browsing this forum: Baidu [Spider] and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
eXTReMe Tracker
Powered by phpBB® Forum Software © phpBB Group
Designed by ST Software.