If the schema in the RFC is authoritative, then we have a further issue, in that the narrative text of the RFC seems pretty clear that only one <rgpStatus> element should be present.
Section 4.1.2 says: The <rgp:infData> element contains a single <rgp:rgpStatus> element... Section 4.2.5 says: The extension element contains a single child element (<upData>) that itself contains a single child element (<rgpStatus>)... None of the examples show more than a single <rgpStatus> element. So I think two things are needed: 1. correct the XSD published by IANA; 2. file an errata with corrected text to allow multiple elements. Ideally, we'd want to add a constraint on the <rgpStatus> element to disallow duplicates (eg two <rgpStatus s="addPeriod"/> elements) but that is probably beyond what an errata can do. Do we need a -bis?! G. > On 20 Feb 2025, at 17:38, Hollenbeck, Scott <shollenb...@verisign.com> wrote: > >> -----Original Message----- >> From: Gavin Brown <gavin.br...@icann.org> >> Sent: Thursday, February 20, 2025 12:15 PM >> To: REGEXT Working Group <regext@ietf.org> >> Cc: Hollenbeck, Scott <shollenb...@verisign.com> >> Subject: [EXTERNAL] RFC 3915 and <rgpStatus> elements >> >> Caution: This email originated from outside the organization. Do not click >> links >> or open attachments unless you recognize the sender and know the content is >> safe. >> >> Greetings, >> >> There is a difference between the XML schema published in the IETF XML >> Registry ([1]) and that inlined into RFC 3915. I can't find any indication >> that >> this has previously been noticed. >> >> Disregarding irrelevant differences in whitespace and comments, the >> difference >> is in the definition of the <rgpStatus> element. In the schema published by >> IANA ([1]), it is: >> >> <element name="rgpStatus" type="rgp:statusType"/> >> >> In the RFC, it is: >> >> <element name="rgpStatus" type="rgp:statusType" >> maxOccurs="unbounded"/> >> >> This means that, depending on where the schema was sourced from, different >> EPP implementations will disagree on how many <rgpStatus> elements can be >> present in EPP commands and responses. >> >> I have no idea how this happened, but I also have no idea how it should be >> fixed. Which version is authoritative? > > [SAH] I don't know how this happened, either, but the RFC is authoritative. > It's possible that a mistake was made when adding the schema to the registry. > >> * Should the XSD in the IETF registry be updated to match the RFC? >> * Should an errata on the RFC be filed? > > [SAH] My preference would be to update the registry to match the RFC. > >> I worry that the first option might have an impact on implementations which >> automatically pull XSD files from the registry. >> >> The RFC itself is vague in its intent. Notwithstanding the >> maxOccurs="unbounded", throughout the text it says that there can only ever >> be a single <rgpStatus> in <info> responses and <update> commands. >> >> How should this be resolved? > > [SAH] I'm interested in hearing other opinions, but as I stated above my > preference is to update the registry. > > Scott -- Gavin Brown Principal Engineer, Global Domains & Strategy Internet Corporation for Assigned Names and Numbers (ICANN) https://www.icann.org _______________________________________________ regext mailing list -- regext@ietf.org To unsubscribe send an email to regext-le...@ietf.org
