[Open Question] Metadata canonicalization — how far to align with ISO 20022 / FIX / FpML

[oa-sarvesh]

Context

The metadata layer (Whitepaper §6.4) proposes a layered model: a common core descriptor, semi-common lifecycle/terms, and issuer-defined extensions, anchored on-chain via content-addressed hashes. It recommends aligning the off-chain record format with established financial data models (FIX/ISO 20022 semantics, FpML for derivatives) rather than inventing parallel vocabularies.

The question

How far should the core descriptor align with ISO 20022 / FIX / FpML, and where does forcing alignment do more harm than good?

How to contribute

• Map core descriptor fields (asset class, issuer DID, ISIN/CUSIP/FIGI, jurisdiction, legal-wrapper reference, behavior manifest) to their nearest ISO 20022 / FIX / FpML equivalents.
• Identify fields where those models have no good equivalent (e.g., capability manifests, settlement-asset trust tier) and native definitions are warranted.
• Flag known pitfalls from prior mapping efforts (e.g., TTF property sets, ISO 20022-to-DLT projects).

Reference: Whitepaper §6.4, §4.5 (TTF).