[urn] Re: 2nd request to register URN for WMO

Anna Milan <[email protected]> Thu, 1 May 2025 14:03:16 +0000
Newsgroups gmane.ietf.urn
Message-ID <PAVPR02MB90632E6A676EE14A6FF61DF6D4822@PAVPR02MB9063.eurprd02.prod.outlook.com>
Dear Dale,
Thank you for your confirmation. We (WMO) understand that center-id and local-identifier SHALL have the necessary requirements for URNs, and we are working on amending the text in the manual to clarify the character limitations.

Please let me know if you need anything else from me.


Best regards, Anna Milan

Scientific Officer

WIS Section, Data and Prediction System (DPS) Division

Infrastructure Department, WMO

https://community.wmo.int/activity-areas/wis

 [cid:ead2262d-f0ac-49d8-aba5-659dcb8fd319]



________________________________
From: Dale R. Worley <[email protected]>
Sent: Thursday, April 24, 2025 21:15
To: Anna Milan <[email protected]>
Cc: [email protected] <[email protected]>; [email protected] <[email protected]>
Subject: Re: [urn] 2nd request to register URN for WMO

My apologies for not responding to this sooner.

> From: Anna Milan <[email protected]> writes:
> Date: Mon, 3 Mar 2025 16:18:30 +0000
>
> [...]
> I'd like to point out that the intent of the request is to have an
> official WMO URN namespace, not to register a complete URN
> schema. Such that, the metadata identifier can be encoded as "urn:wmo"
> instead of the previous practice of "urn:wmo-x".  Is it possible to
> just register a WMO URN namespace without the schema details or can
> you provide information about another mechanism for this request?
> [...]

In terms of the IETF terminology, there are "informal" and "formal" URN
namespaces.  Informal namespaces are assigned with minimum registration
information, but are namespaces like "urn-1", "urn-42", and "urn-2025".

Formal namespaces require some sort of clear definition.  However, RFC
8141, section 6.3 notes that the process is simpler for "Standards
Development Organizations, Scientific Societies, and Similar Bodies": "A
specification is required" but that only needs to "point to the needed
external standards or specifications."

In the case of urn:wmo, all that is needed is to provide a good
reference to the WMO's definition, as long as that definition provides
the needed information.

> From: Anna Milan <[email protected]>
> Date: Wed, 23 Apr 2025 08:26:44 +0000
>
> Thank you for your feedback. Please see the revised request. As
> mentioned before, we are requesting to only register WMO namespace of
> which urn:wmo:md is one implementation. I hope the below captures this
> intent sufficiently.

> The syntax of wmo NID-based URNs has the following structure:
>
> urn:wmo:{resource-type}:{centre-id}:{local-id}
>
> Where:
>
> Currently, only the “md” NSS is defined.

Of course, that should be 'the "md" resource-type'.

>   The requirements and syntax for discovery metadata identifiers are
>   specified by the WMO Core Metadata Profile (Version 2), available in
>   the Manual on the WMO Information System (WMO-No. 1060), Volume II -
>   Appendix F (https://eur01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flibrary.wmo.int%2Fidurl%2F4%2F68731&data=05%7C02%7Camilan%40wmo.int%7Cf95616d7f7414edfb6e708dd83645bd2%7Ceaa6be54468740c49827c044bd8e8d3c%7C0%7C0%7C638811189262492510%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=n5HbLYY2MCZXwSeI0SeWdaSfUSF7Tq%2BLvHTAZznIrCY%3D&reserved=0<https://library.wmo.int/idurl/4/68731>).  Future NSS values
>   will be defined as required by WMO.

Very good.

> The "centre-id" is managed by the WMO Secretariat.  Codes are proposed
> and approved during the WMO Member registration process for data,
> metadata, and service publications.

It looks like the governing standard for "centre-id" is Appendix F
section 1.4, requirement 2, clause C, which points to Appendix D section
1.6.

Everything that section says is permitted is acceptable in URNs.  In
particular, dashes are expected to be used.  It also says that "The
centre identifier (centre-id) is an acronym, as specified by the Member
and endorsed by WMO Secretariat."  Presumably "acronym" implies a
requirement that it consists of letters (and dashes).  It needs to be
understood that the acronym must not contain accented letters, which is
not stated.  Though it is likely that the Secretariat that approves
center-id's is already enforcing this constraint.

> The "local-identifier" is managed by the WMO Member.

The governing standard seems to be Appendix F, section 1.6, requirement
2, clause D.  In particular "The local identifier SHALL NOT have spaces
or accented characters."  The implication is that the local identifier
can contain unaccented letters (presumably meaning the 26 "basic Latin"
characters).  And I expect that digits are allowed.  But the text does
not explicitly limit the punctuation characters that might be allowed.
Does the WMO's practice/understanding exclude punctuation characters
which are not permitted in URNs?

I note that in that section is permission 1, which explicitly allows
colon characters in local_identifier.  But that is OK for WMO URNs, as
the syntax of the "id" property given in the section shows that a colon
in the local_identifier does not cause any ambiguity in parsing it.

If we have assurance that the WMO understands that center-id and
local-identifier have the necessary requirements, I agree with accepting
this registration.

Dale

_______________________________________________
urn mailing list -- [email protected]
To unsubscribe send an email to [email protected]
Outlook-mwkh11mh.png (image/png, 35 KB) - not displayed