Skip to main content

2026-01-23 1154 AEDT

Jan 23, 2026

UN CEFACT GTR - AEST / PST​

Invited Jo Spencer anica@registradores.org Steve Capell John Phillips

Attachments UN CEFACT GTR - AEST / PST

Meeting records Transcript Recording

Summary​

John Phillips, Sankashan, Steve Capell, and Ann Dao discussed the UN/CEFACT global trust registry project, focusing on the need for stable documentation and a publishing strategy that aligns with UN/CEFACT governance distinctions between technical standards (requiring only bureau approval) and recommendations to member states (requiring plenary approval). Key topics included clarifying that The Grid and DIA should be communicated as containers for existing identifiers, the importance of an issuing authority only signing for data they control within the DIA, and the decision to use the term "UN Member State" over "nation state" in project documents. John Phillips and Sankashan offered guidance to Ann Dao on navigating GitLab, and Sankashan reported that Brett Highland's feedback was converted into ten specific issues for resolution, with a plan to have documents ready for public review by the end of March.

Details​

  • Project Context and Rules John Phillips opened the UN/CEFACT global trust registry project meeting, reminding participants that it is a UN/CEFACT project operating under rules like the open development process and royalty-free public accessibility. The meeting agenda included reflecting on previous discussions and working through GitLab issues, with Sankashan and Steve Capell contributing content [00:00:00].

  • Project Recap and Identifier Schemes John Phillips recapped discussions from two weeks prior, noting interest from His Majesty's Revenue and Customs from the UK, whom they intend to re-engage. The discussion also covered the European identifier as part of the EIDAS framework, emphasizing that the Grid and DIA should be communicated as a container-type structure, not competitors, to existing national, international, or regional identifier schemes [00:01:05].

  • Privacy and Consent John Phillips shared that Mark Lazar, an expert in consent and consent receipts, provided new content on the privacy aspects of content. John Phillips personally felt that the Grid and DIA do not yet need to incorporate the concept of individual content, although they may (will likely) need to reach that point eventually. [00:02:01].

  • Project Timelines and Publishing Strategy John Phillips clarified that the UN/CEFACT plenary is scheduled for November, not July or May as previously thought, but Steve Capell advised that the project should focus on producing "useful, stable, good documentation" and suggestions for next phases, which can be published when ready, regardless of formal meeting dates [00:02:57]. Steve Capell emphasized that the most important thing is to publish quality, tested material that has reached a level of consensus [00:04:48].

  • UN/CEFACT Governance and Document Approval Steve Capell explained that the UN CFAC governance distinguishes between technical standards, such as a good part of the Grid and the DIA, which only require bureau approval, and recommendations to member states, which must be plenary approved. Technical standards, once ready, can proceed to public consultation, a 30-day window following approval by the bureau [00:06:08].

  • Recommendations for Member States Steve Capell suggested that the project is a candidate for producing both a technical specification and a plenary-approved policy recommendation to member states on why they should implement the Grid. To meet the November plenary deadline, a draft policy recommendation would need to be produced around the end of May to accommodate translation and a two-month public review period [00:07:10]. John Phillips was encouraged by this timeline, suggesting that the specification documents could be ready for publishing by the May forum, positioning them to draft the recommendation afterward [00:08:21].

  • Extension of the Pattern for Non-UN Contexts Steve Capell raised an issue concerning the need for specificity and verification of registry membership in spaces like the UNP project, specifically for product and facility IDs, which may not currently fit within the Grid's business scope. Steve Capell explained that the pattern used by the Grid—a register attesting membership with an identity credential, and an international club recognizing the register—repeats in other contexts not governed by the UN, such as conformity assessment accreditation [00:09:07].

  • Globally Unique Identifiers for Registers Steve Capell proposed taking the same pattern and applying it to facilities and products within UNP but outside Grid governance, requiring extra data elements for registers to declare their scheme ID and a scheme pattern for generating a globally unique ID for a given member [00:10:29]. This pattern addresses the argument that national business numbers are not globally unique by suggesting concatenation of the register's globally unique identifier with the member's identifier to form a URI [00:11:42].

  • Technical Endpoint and Data Expression John Phillips noted that the proposed changes from Steve Capell's merge request effectively replace a vague "technical endpoint" line with multiple lines that better express what a technical endpoint might be. John Phillips further noted that a concatenation opportunity exists within the Grid using a country's unique ISO three-alpha code prepended to a register's information to ensure a unique string, since registers already issue unique identifiers [00:12:49]. Steve Capell differentiated between the URL pattern for a human-viewable page, which might be a consequence of technical implementation, and the authoritative globally unique URI or URN for a member, necessitating separate lines for these items [00:13:54].

  • Review of Merge Request and Feedback John Phillips, referencing yesterday's discussion that Steve Capell was not part of, indicated that there had been a concern that the repeatable pattern might include responsibilities or scope within the Grid, but John Phillips clarified this was not the intent [00:14:57]. John Phillips planned to leave the issue open until Monday morning for others to review, expecting it to be merged [00:15:46].

  • Actioning Brett Highland's Feedback Sankashan reported that they had converted an email of feedback from Brett Highland into 10 specific, linked issues for review and fixing on the website [00:15:46]. Sankashan added scope and acceptance criteria to each linked item to streamline the review process [00:16:51].

  • Diagrammatic and Terminological Changes John Phillips acknowledged that one of the observations from Brett Highland was the repeated use of the same diagram across different contexts on the website, suggesting the need for different diagrams to highlight specific areas of the framework [00:18:06]. John Phillips also mentioned addressing a query from a professor about the use of the term "nation state," noting that the term means more than initially thought [00:19:02].

  • Use of "UN Member State" Terminology John Phillips proposed changing the term "nation state" in project documents to "UN member state" because it is more specific and accurate for the intended use within the Grid, based on AI recommendations [00:19:02]. Steve Capell supported moving away from "nation state" and "economy," suggesting "UN member state" for governance contexts and "country" for general talk about nations, although John Phillips pointed out complexities with "country" related to First Nations people in Australia [00:20:46]. The consensus leaned towards "UN member state" due to its legal significance [00:21:43].

  • Developing the DIA Spec John Phillips introduced a chunky issue regarding what changes, if any, are needed for the existing DIA specification to make it more powerful or fit within the Grid model [00:21:43]. John Phillips initially promoted the idea of distinguishing between identifiers issued by the registrar (under their control) and DIDs supplied by the applicant (asserted and verified by the registrar), suggesting two types of data in the DIA [00:22:49].

  • Concerns about Data Explosion in DIA The previous day's discussion, including Harmon van der Kooij, raised concerns about risking an "explosion of functionality" inside the DIA. John Phillips noted that the idea was not to mandate usage but allow it [00:23:49]. Steve Capell agreed, viewing this as a classic data explosion challenge, advising against including data in the DIA that the authority does not control, especially non-verified data [00:24:45].

  • Authority Control and DID Verification Steve Capell emphasized that the authority issuing the DIA should only sign for data they control. They stressed that a DID can only be included if the authority verifies control of the DID by the applicant before issuing the DIA, making it an attestation of a point-in-time verification [00:24:45] [00:27:54]. Steve Capell suggested that any extension points in a DIA should be for the authority to choose what extra data they want to include, not for the recipient to dictate random data [00:25:56].

  • Optionality for Identifiers in DIA John Phillips recounted a conversation about the KVK (Dutch Ministry of Commerce) issuing three types of identifiers—KVK number, LEI, and future EU ID—and whether they should issue three separate DAOs or one DIA with three identifiers [00:28:57]. John Phillips' stance was to provide KVK with the choice [00:29:54]. Steve Capell noted that while having three identifiers in one DIA might be feasible if the authority controls and verifies them, there is a risk of complexity, especially regarding the different legal texts that support the issuance of each identifier [00:30:45].

  • Focusing on Trust Anchoring in DIA Both John Phillips and Steve Capell agreed that the purpose of the DIA is trust anchoring, and it should be kept "lean and mean" [00:28:57]. Steve Capell suggested it is reasonable and feasible to separate different identifiers and signing processes (e.g., using DIA for KVK or vLEI for GLEIF) rather than confusing them within one DIA [00:32:46].

  • Ann Dao's Integration and Questions Ann Dao joined the discussion, expressing the need to learn how to navigate GitLab. John Phillips and Sankashan explained that Ann Dao could create an issue on GitLab or propose changes directly via the "edit this page" function, which should generate a merge request [00:34:42]. Ann Dao asked about the conclusion regarding the UNP DIA changes, and John Phillips summarized that the momentum is shifting away from including multiple data types beyond those controlled by the registrar and the DIDs supplied and verified by the applicant, due to the risk of complexity [00:37:25].

  • Managing Iterative Issue Updates Sankashan advised John Phillips on how to manage the evolving nature of issues, suggesting that if the discussion renders a baseline issue summary redundant, they should open a new issue with the updates and resolve the old one with a closing comment. John Phillips considered creating a draft submission to the UNP spec within their GitLab environment, treating it like a document, to manage versions and gather feedback [00:40:03]. Sankashan viewed this approach as a legitimate "project fork" to build and package changes before sending a merge request to the UNP upstream [00:42:30].

  • Public Review Process Ann Dao asked about the public feedback process, and John Phillips explained that the technical specification needs a period of public review, which formally begins when the documents are submitted to the UNCFAC bureau as stable and complete [00:45:36]. This is distinct from the prescribed public review period required for a formal recommendation to member states [00:46:45]. The goal is to have the documents ready for public review by the end of March to allow time for handling comments and making necessary changes before the May forum [00:47:45].

  • Document Consolidation for Review John Phillips indicated that the public review could be conducted on the existing separate documents in GitLab because it is easier to manage comments and feedback. They suggested that a single PDF or printable version, consolidating the separate documents, could be produced after the public review is complete, without changing the content [00:48:58]. John Phillips also noted that after public review, the content will likely need to be formatted to fit the UN/CEFACT fact standard, which is why they do not want to do this work too early [00:48:58].

Suggested next steps​

  • John Phillips will work to finalize and publish the project documents, aiming for the end of March.

  • John Phillips will leave the issue open until Monday morning to allow others to read and review it, and then proceed with acceptance.

  • John Phillips will go through the project documents and change the use of 'nation state' to 'UN member state' where appropriate for the grid's intended use.

  • sankarshan will consider and update the issue with the best way to handle the conversation about the UNP DIA spec, such as creating a new document or a project fork.

Chat​

00:34:26

steve capell: my dear colleagues, I have to cross town for another meeting. I hope you dont mind if I sign off now.

00:35:10

John Phillips: https://opensource.unicc.org/un/unece/uncefact/gtr/-/issues