Update Authenticator Definition #1195
Conversation
|
Is there an issue corresponding to this PR? |
|
Mostly #1175 - I wanted to draw more attention to the fact that the authenticator could be defined in either software or hardware |
|
Thanks for the edits! I think this is much more concise 👍 |
|
The current definition doesn't mention that authenticators can be grounded in software, which I think is something important to denote. I think there is confusion currently around what an authenticator is, and this is a helpful point to make. |
Current wording includes “or a software component of the [=client=]”. |
|
I believe the wording that @agl quotes does the trick. |
|
It isn't only "of the [=client=]" though, it could also be a component of the client device. |
s/ownership/possession/
|
@akshayku can you approve that the changes from your comments have been added? |
index.bs
Outdated
| cryptographically signing and returning, in the form of an [=Authentication Assertion=], | ||
| a challenge and other data presented by a [=[WRP]=] (in concert with the [=[WAC]=]). | ||
| :: A cryptographic entity, existing in hardware or software, that can [=registration|register=] a user with a given [=[RP]=] | ||
| and later [=Authentication Assertion|assert possession=] of the registered [=public key credential=] when requested by the [=[RP]=]. |
There was a problem hiding this comment.
I don't think we want to loose the notion of user verification:
suggest:
...assert possession=] of the registered [=public key credential=], and optionally [=user verification|verifying the user=], when requested by the [=[RP]=].
|
@nicksteele Please see @equalsJeffH comments, and update so we can merge this |
|
Changes have been made and are ready for review @equalsJeffH |
Preview | Diff