SourceForge : artf6573: Make Location a subclass of Network Object

Project: NML-WG Trackers > Schema Progress > View Artifact

Artifact artf6573 : Make Location a subclass of Network Object

Tracker:	Schema Progress
Title:	Make Location a subclass of Network Object
Description:	Proposal: Make Location a subclass of Network Object Rationale: 1. Location has an id and name attribute. These attributes are described Network Object, so making Location a subclass of Network Object prevents us from defining an id and name specific to the Location. 2. The 'locatedAt' relation relates a Network Object to a Location. I can imagine a future extension where NML allows Locations within a Location (e.g. one Location describing a country, another Location describing a city, a third Location describing a specific server room). If Location is a subclass of Network Object, such relations may be allowed in the future. Otherwise, it is not possible to make such extension.
Submitted By:	Freek Dijkstra
Submitted On:	07/18/2012 6:19 AM EDT
Last Modified:	11/30/2012 7:53 AM EST
Closed:	11/30/2012 7:53 AM EST

	Status / Comments		Change Log		Associations		Attachments

Status
Status:*	Rejected
Category:*	– Visualization / Geo-data
Priority: *	2
Assigned To: *	None

Comments

Jeroen van der Ham: 11/30/2012 7:53 AM EST
	Action:	Update Closed set to 11/30/2012 Status changed from Last Call - Reject to Rejected

Freek Dijkstra: 09/06/2012 9:46 AM EDT
	Comment:	There seems consensus on rejecting this idea and either re-define the "id" and "name" for the Location, or in some base object (like AnyElement in XML or Resource in RDF or DataObject in OWL).
	Action:	Update Status changed from Under discussion to Last Call - Reject

Jason Zurawski: 08/16/2012 7:16 AM EDT

Comment:

I am having a hard time following this one (probably because I wasn't paying attention or attending a call).  Can some examples be encoded to show the
 plusses and minuses?  Given its relationship to another open/soon to be closed issue, maybe this is all moot?

Action:

Update

Freek Dijkstra: 08/09/2012 11:55 AM EDT
	Comment:	too mature --> too premature. (doh)
	Action:	Update

Freek Dijkstra: 08/09/2012 11:54 AM EDT

Comment:

There is discussion, so last call was probably too mature.

Jeroen -- OK, let's reject this proposal.

How about simply allow the "id" and "name" to be generic so they can be used for all resources, not just network resources? I think this is possible 
in both RDF and XML schemas.

Action:

Update
Status changed from Last Call to Under discussion

Jeroen van der Ham: 08/09/2012 7:13 AM EDT

Comment:

How about just adding "id" or "name" as an attribute to the Location?

All current NetworkObjects have some kind of role in the processing and forwarding of data. Locations have nothing to do with that. It seems a bit 
short-sighted to make Locaton a Network Object just to make it have a name and id. Other than that it serves no function.

Action:

Update

Freek Dijkstra: 08/09/2012 5:48 AM EDT

Comment:

Jeroen, you are right, but forget about the first reason in the description. In the current schema it is invalid to identify a Location by a URN, 
because it has no id or name. We need to fix that, and making Location a subclass of NetwObj seems the easiest fix.

Action:

Update

Jeroen van der Ham: 08/09/2012 5:33 AM EDT
	Comment:	Now that we have a proposal for Relations to no longer require NetworkObject to NetworkObject (https://forge.ogf.org/sf/go/artf6563), I see no reason to make this change.
	Action:	Update

Freek Dijkstra: 08/08/2012 10:26 AM EDT
	Action:	Update Priority changed from 4 to 2 Status changed from Under discussion to Last Call

Freek Dijkstra: 07/18/2012 6:19 AM EDT
	Action:	Create

Return