- From: Mark Nottingham <mnot@mnot.net>
- Date: Wed, 7 May 2025 11:21:00 +1000
- To: Matt Peterson <Matt.Peterson=40entrust.com@dmarc.ietf.org>
- Cc: "Julian F. Reschke" <julian.reschke@greenbytes.de>, "ietf-http-wg@w3.org" <ietf-http-wg@w3.org>, "draft-ietf-scim-cursor-pagination.all@ietf.org" <draft-ietf-scim-cursor-pagination.all@ietf.org>, "last-call@ietf.org" <last-call@ietf.org>, "scim@ietf.org" <scim@ietf.org>
Hi Matt, > On 7 May 2025, at 2:09 am, Matt Peterson <Matt.Peterson=40entrust.com@dmarc.ietf.org> wrote: > > Using this criterion, should we also list TCP, and IP as normative references? It seems overly complicated for every specification to reference the flattened list of the entire tree of normative references. This will in most cases result in a *very* long list of normative references which makes it difficult to focus the reader’s attention on the few specification(s) that are critically useful in implementing the specification. No; RFC9110 normatively references TCP, and the reference is transitive. > In the case of SCIM Cursor-based pagination, the “nearest” and most relevant normative reference is RFC 7644. THIS is the spec that implementers must understand.... and if they need to also understand HTTP, then they can follow RFC 7644’s normative references... and the normative references in those RFCs until they get to a level (like TCP, IP) where understanding the details is irrelevant. That makes sense - I see that 7644 normatively references HTTP, so normatively referencing 7644 should be enough. Cheers, -- Mark Nottingham https://www.mnot.net/
Received on Wednesday, 7 May 2025 01:21:11 UTC