Add notes for validFrom and validUntil.

[msporny]

Reserve validFrom and validUntil, note expected changes in next version of spec.

Related to #584.

---

Preview | Diff

[dlongley]

I think there's a slight problem here for JWT representation. We current express issuanceDate as the "iat" claim in a JWT. The validFrom property should instead map to the "nbf" claim in a JWT.

It is true that we intended "issuanceDate" to mean validFrom in the absence of that property. However, I think both properties could be present in a VC with different values and, under that circumstance, validFrom should take precedence for determining validity. So issuanceDate means the earliest point at which a VC should be considered valid, unless validFrom is present. Of course, we can't say that normatively now and we're just reserving this for the future.

So I think it's still true that the range of acceptable values will remain the same and the semantic meaning of the property will be the same but this will not be a "renaming", rather, the presence of validFrom would take precedence over issuanceDate when determining the VC validity period. Also, as mentioned above, mapping these properties to a JWT would result in using different JWT claims.

cc: @TallTed

[TallTed]

Good catch. I think that in the current iteration, we must express issuanceDate as nbf (as this is and was the intended semantic), and not as iat (which was never the intended semantic). Future versions of the VCDM can change things however is appropriate at that point, including how precedence lays out if/when both are present.

[msporny]

Ok, updating the PR w/ these changes in mind.

[msporny]

Fixed in afaedca

[msporny]

Need a re-review on this one as it too complicated to just merge on its own.

[msporny]

@dlongley and @TallTed, can both of you provide a re-review please?

[TallTed]

@msporny, @dlongley -- I think associated changes (which I think would be best made in the same merge) are also needed to the JWT mapping section, particularly (in the msporny-issue-584 branch) the following, which are now focused on JWT exp and iat and do not mention nbf --

• lines 3443-3444
• lines 3450-3451
• lines 3524-3527
• lines 3534-3536

[TallTed]

As it seems my review is the blocker here...

I cannot approve this without the changes to the associated areas that I highlighted previously.

I think these would count as bugfix, though they would change normative text, and should therefore not trigger a new CR.

[stonematt]

PR was merged to make way for the resolution from WG call on 18 June 19:
RESOLVED: Change spec text mapping of issuanceDate to use JWT nbf claim (with matching semantics) instead of JWT iat claim (with conflicting semantics) (#646), in parallel to testsuite change to match (w3c/vc-test-suite#50)