AuthorizedKeyFrag and assorted fixes#2745
Conversation
KPrasch
left a comment
There was a problem hiding this comment.
Approved with comments and one renaming RFC.
3671df3 to
3d2103d
Compare
e759510 to
b47b2a8
Compare
derekpierre
left a comment
There was a problem hiding this comment.
Any reason for the empty newsfragment file?
| response_data = {'treasure_map': new_policy.treasure_map, | ||
| 'policy_encrypting_key': new_policy.public_key, | ||
| 'alice_verifying_key': new_policy.alice_verifying_key} | ||
| # TODO: technically we should return both Alice's and Publisher's keys here. |
There was a problem hiding this comment.
If not being addressed in the PR, does this need a follow-up issue?
There was a problem hiding this comment.
I am actually not sure now if the interface warrants it. Can a policy with Provider and Publisher being different characters be even created at the moment, in any way?
There was a problem hiding this comment.
After some thought, I think it's justified to return only the Alice's key - currently for this interface there's no way to have a separate publisher. When this capability is added, we can extend the return value.
Initially it only contained internal changes, but I guess now the Alice/publisher stuff has leaked into the public API. I'll add one. |
b47b2a8 to
5786c59
Compare
61fda47 to
2ffcfd8
Compare
2ffcfd8 to
8c628da
Compare
|
|
||
| try: | ||
| verified_kfrag = kfrag.verify(verifying_pk=alice.stamp.as_umbral_pubkey()) | ||
| verified_kfrag = authorized_kfrag.kfrag.verify(verifying_pk=author.stamp.as_umbral_pubkey()) |
There was a problem hiding this comment.
Nice! That reads a lot better. 🔥
Type of PR:
Required reviews:
What this does:
Bob.retrieve()and into the respective interfaceAuthorizedKeyFragclassSee #2752 regarding the renaming.
One thing not changed in this PR (should it be?) is that the
grantinterface returns only one verifying key (still namedalice_verifying_key). Technically it should return both Alice's and publisher's key, but using web interface implies that Alice is the publisher, so perhaps it's ok?