We discussed how the tree of possibilities in KERI can't cloud the simplest steps. And so the spec needs to be very clear about what is optional/additional and needs to logically build in a way that makes it obvious to the reader where they can stop for their use case.
Catchup on final changes status:
- What needs to be completed to enter the public review process?
- All public review changes should be well documented so that they can watch the process
- Normative vs Informative information in the spec is important prior to release https://github.com/trustoverip/tswg-did-method-webs-specification/issues/120
- Work backwards from the normative statements. Triage informative to minimum.
- If you start with these bones it's easier to hang all of the minimal informative information and move everything else outside of the spec.
- Informative that isn't needed in the spec
- Appendix
- Implementation guide
- Explain why in a separate paper
Breakdown by section:
See the ToIP template, might have ISO influence
- Normative
- Core Characteristics - 'This section is normative' (see examples in other specs)
- Method Name
- 'webs' SHOULD is a normative but for an informative
- Method-Specific Identifer ABNF is Normative, see DID core spec
- KERI AID paragraph contains informative statements, which is the opposite of the previous Normative section. Good
- Should some of those statement be Normative?
- And if they are altered to be Normative, should it be referencing the normative info in the KERI s pecs?
- Breakdown normative into addressable numbered sections that provide the rules to build from.
- You cannot include a non-Normative reference in W3C
- if SCID has to be AID, then AID must be normative.... so perhaps specify a SCID... bare minimum is preferred.
- Informative
- Abstract (shouldn't be numbered)... it is simply informative
- Introduction
- This is a specific section currently, but probably should become all of the informative information at the front of the spec
- Requirements, Notation, and Conventions
- Informative inside normative should be boxed... but only if it helps devs etc make a decision
- Use examples at the start of these informative sections
- Todos and missing information sections need to be addressed today
- Reivew current state of hyperledger-labs did:webs reference implementation https://github.com/hyperledger-labs/did-webs-resolver
- Open PRs and Issues