-
Notifications
You must be signed in to change notification settings - Fork 35
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
fix: cannot set salts for credential creation (suggestion) #222
fix: cannot set salts for credential creation (suggestion) #222
Conversation
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## development #222 +/- ##
===============================================
- Coverage 82.86% 82.85% -0.01%
===============================================
Files 47 47
Lines 4202 4200 -2
Branches 1047 1046 -1
===============================================
- Hits 3482 3480 -2
Misses 691 691
Partials 29 29 ☔ View full report in Codecov by Sentry. |
It's good if the API can accept exn messages directly, without property mapping. The API probably becomes harder to use with the compact one letter property names. Need to make sure all required properties are declared and have some reasonable documentation. |
Signed-off-by: Nuttawut Kongsuwan <[email protected]>
schemaId: SCHEMA_SAID, | ||
recipient: holderAid['prefix'], | ||
data: { | ||
await issueCredential(client3, 'issuer', { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The readability of code is paramount.
We should fix this at the serialization level.
One option:
https://www.npmjs.com/package/typescript-json-serializer
@JsonObject()
export class Animal extends LivingBeing {
@JsonProperty() name: string | undefined;
@JsonProperty() birthDate: Date;
@JsonProperty() numberOfPaws: number;
@JsonProperty() gender: Gender;
// Enum value (string)
@JsonProperty() status: Status;
// Specify the property name of json property if needed
@JsonProperty('childrenIdentifiers')
childrenIds: Array<number>;
constructor(name: string | undefined) {
super();
this.name = name;
}
}
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This uses decorators on the objects. And leaves the code readable.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
In my opinion, decorators for serialization in JavaScript do not add any value. If we want to keep the API with more descriptive property names, I would vote for reverting back to plain property mapping. But my preference would be to move towards less mapping rather than more. I.e., using the same object shapes as KERIA accepts and returns.
I just listened in on the dev call and wanted to provide some comments. It seems like this change sparked some discussion on which direction to take regarding the API. I am voting for going towards a signify API that as closely as possible resembles the API of KERIA. The additional property mapping just is just another layer of indirection that in itself requires documentation. Some benefits of the approach:
Also, for this particular API of issuing credential, a user of this library could add their own functions for ACDC creation according to which ever schema they want to use, like this. function createQviCredentialData(args: QviCredentialData): CredentialData {
return {
ri: args.registryId,
i: args.issuerPrefix,
s: "EBf...",
a: { LEI: args.LEI, i: args.recipientPrefix }
}
}
const data = createQviCredentialData({ registryId, issuerPrefix: "abc", recipientPrefix: "def", LEI: "123" })
await credentials().issue("issuer", data) As @psteniusubi mentioned as well, the fact that the API can accept the same shape of objects that is passed in exchange messages is a huge plus in my opinion, as can be seen in this change: 7022f95#diff-6d0716ef2bd3e3e9494efe4b175ef940675ac2c0615c4d98172c7c5f69671c9aR906. If we decide to go for this approach, we should probably plan to change the rest of the API to do follow the same strategy and then strictly adhere to it going forward. I do see the argument against it as well though, using more descriptive property names is helpful, but the benefits mentioned are just much greater in my opinion. |
@lenkan - I agree with the suggestion to move forward with compact labels. The Keri spec already has definitions for message types and labels. (I could not immediately find all message types, for example "iss" and "vcp" are missing.) As you suggest it is relatively easy to build a higher lever abstraction on top of the lower level Signify API. One could even implement a backwards compatibility layer. There is at least one catch. Because Keri sets strict rules on ordering of the fields, the Signify API should still internally re-build all messages. I don't think the API can internally use the spread Should this design change apply to signifypy as well? |
51e4971
to
49cd8d5
Compare
…t#222) * fix: cannot set different salts for credential * fix types * add comments * fix dt for issuance event
This is my suggestion on how to solve #221 .
I know this is a breaking change, but I don't think we need the additional interface for IssueCredentialArgs. I think it just adds confusion with the additional mapping of properties. Same for the other operations such as create identifier, rotate etc.