Closed kimdhamilton closed 2 years ago
NB @David-Chadwick
I've been trying to understand Issue # 2, and i am a little stumped. Part of me thinks we should spill a paragraph's worth of free digital ink after the second graf of ## Input Evaluation
(L 1191 in the PR), explaining that the choices taken by processors in Input Evaluation may be judged or tracked by Verifiers, and to consider optional features subject to reputation-scoring. Was this what Daniel meant? Would a mentioned in the definition of Features in the ## Terminology
section help, since they are free from conformance-stick but still subject to the reputation-carrot?
Sidenote, are these 5 issues being opened as tracking issues if not addressed by this merge? Not sure if that was implied or not but asking for the sake of clear process.
Sidenote, are these 5 issues being opened as tracking issues if not addressed by this merge? Not sure if that was implied or not but asking for the sake of clear process.
It is a mix of "notes to self" that I will address before merging (e.g. index.html hack), issues we may quickly decide on as a group (and I can address with this change", and (default) issues I will create as separate tracking issues if not address.
I've been trying to understand Issue # 2, and i am a little stumped. Part of me thinks we should spill a paragraph's worth of free digital ink after the second graf of
## Input Evaluation
(L 1191 in the PR), explaining that the choices taken by processors in Input Evaluation may be judged or tracked by Verifiers, and to consider optional features subject to reputation-scoring. Was this what Daniel meant? Would a mentioned in the definition of Features in the## Terminology
section help, since they are free from conformance-stick but still subject to the reputation-carrot?
Definitely that's the right line of thinking -- this is just an area I hadn't already addressed in the PR but something I want tracked. So worst case, if I don't have time to get to it in this already massive change, we can create a tracking issue
Based on discussion on Jan 27th:
Note # 1: limit_disclosure
should be moved into the base profile with additional language around "required"
("If you don't support "required" don't return anything" / "Wallets must understand how to at least know when "limit_disclosure" is "required"")
Note # 2: Have the "status" constraint be it's own feature, the others are relational within the credential but status is about the credential itself
Note # 3:
Note # 4: Having a feature impact multiple objects doesn't feel weird to us
Thanks for the feedback!
I agree with the discussion point that we should land this and follow up with subsequent PRs. Here are next steps I will take
limit_disclosure
was confusing, and now also needs to capture the "wallet doesn't understand" case. The latter is described in Input Descriptors, but needs to be updated here too
See preview
Reflects discussion of Jan 20 Discussion.
Notes
limit_disclosure
is now entirely moved to a FeatureIssues
descriptor_map
? Consider doing this in separate PRFeel free to refer to Note # or Issue # when commenting. E.g., "Note 2: I disagree"
Fixes #283; Overrides #275