Project

General

Profile

CIM Issues #5963

mRID topics [incl GMDM #19]

Added by Yang Feng almost 2 years ago. Updated about 2 months ago.

Status:
Review
Priority:
High
Solution Version:
Breaking Change:
Yes
Breaking Change Description:
May break vendor's implementation
CIM Impacted Groups:
WG13, WG14, WG16
Requestor:
Martin Miller
Standard(s):
Version:
Clause:
Sub-Clause:
Paragraph:
Table:
Origination Date:
Origination ID:
Originally Assigned To:

Description

The usage of mRID has been brought up during the GMDM IOP and followings are the topics to be covered

• Should the UML be changed to make the mRID a UUID instead of a string?
• The omission of mRID in favor of rdf:ID/rdf:about has been inadequately canonized. How can we canonize this in a way that does not retroactively break anyone’s software?
• When the mRID is included, what should be done when the mRID value does not exactly match the fully-expanded rdf:ID/rdf:about?
• What should be done for mRIDs that are not UUIDs but that do not correctly override the xml:base?
• What should be done when a 61968-100 mRID does not conform to the xml:base nor exactly match the fully-expanded rdf:ID/rdf:about?


Files


Decision

Reviewed on 15-Jun-2023 in Oslo:

Decisions:

1. Change the description of mRID to:

Master resource identifier issued by a model authority. The mRID is unique within an exchange context.
Global uniqueness is easily achieved by using a UUID, as specified in IETF RFC 4122, for the mRID. The use of
UUID is strongly recommended.

2. The rdf:ID/rdf:about will be a 552 issue & JSON-LD @ID will be a 553 issue. This will not be tied to the UML. In the 552 rdf:ID/rdf:about is to be decoupled from the mRID attribute and explicitly prescribed to be a UUID. The same being true in the 553. This will allow both serializations to have the same identifier, generated by the same rules to produce the same IRIs. The legacy formats will be based on local business rules instead of standards-based rules.

3. We currently have in Section 8.7 in Part 100 Ed 2.0 the new ObjectIdentificationRevisionSets profile (i.e. for object registries). We need to introduce a new similar profile (RDF-based) and include in the 552. See if we can utilize the existing profile to generate the RDFS or if there needs to be additional work done to the profile. To clarify, this bullet point addresses the identifier now tied to serialization. (Additionally reviewed on 05-Jul-2023 and the below sub bullet points have been added for clarification).

- The existing documentation (for the latest (Names, NameType and NameTypeAuthority) proposal has been forwarded to Chavdar. This is what we agreed needs to be added to the 301 Ed. 8.0.
- Then we need to have a discussion with respect to whether an additional profile is necessary and which would be included in the 552.
- Svein has suggested including currently parked (or some variation) of the draft of the "Naming Service Specification" (i.e. IEC 61970-454). Would need to be revisited and collectively agreed upon but this may be one approach. If so the sugggestion is to include in the IEC 61970-452

#2

Updated by Jim Waight over 1 year ago

WG16 inputs from Alvaro

#3

Updated by Alvaro Marciel over 1 year ago

ENTSO-E position on mRID

#4

Updated by Todd Viegut about 1 year ago

  • Decision updated (diff)
#5

Updated by Todd Viegut about 1 year ago

  • Status changed from New to Open
#6

Updated by Todd Viegut about 1 year ago

  • Status changed from Open to Review
#7

Updated by Todd Viegut about 1 year ago

  • Decision updated (diff)
#8

Updated by Todd Viegut about 1 year ago

  • Decision updated (diff)
#9

Updated by Todd Viegut about 1 year ago

  • Decision updated (diff)
#10

Updated by Becky Iverson 7 months ago

Testing making updates to the Redmine issue.

Also available in: Atom PDF