Closed loneil closed 1 month ago
I'm not sure we support did:peer:1 in ACA-Py, but not certain. @dbluhm @Jsyro, can you please remind me? We don’t need to emit a DID:peer:1, but it would be good to be able to receive one in a DIDComm connection.
Next issue is the verification of the DID types we support. This is the second time this week this issue has come (Issue 2714 in ACA-Py. For now, lets track them down as we hit them and get the correct verification check in place. Likely this is doing a “indyDID” validation instead of “supportedDID” verification — something like that.
Nice to track down where the ACA-Py problem is and add an ACA-Py issue.
Also — we want BC Wallet to use (ideally) peer:did:4 — I’ll see if that is feasible.
@swcurran We support resolving did:peer:1 that we've received in a did exchange now. That's in main but not in a release yet.
Awesome - thanks @dbluhm — good to know. I think we need to push a release out. Lots in the pipeline.
That does mean we still have to eliminate the overly aggressive validation check. @loneil — great if you could track down where that is happening and raise the issue in the ACA-Py repo.
@swcurran I just ran a quick test; on 0.11.0, the overly aggressive validation is present (as we see here in this report). On nightly, we passed the validation and it was actually the version of AFJ I was testing against that failed on processing our response from ACA-Py. So I think this is resolved in 0.12.0rc0.
Ah great, thanks for the info @dbluhm . I'm going to try tests with 0.12.0rc0 when I have a moment so can try this out with that.
This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 5 days.
Testing this in ACA-Py 0.12.0rc2 I no longer run into ValueError: Bad specification did:peer:1zQmR4vLev39FVRRWfoJQp9BjfEu1CuHnxo69qLqwPnE4jJJ does not correspond to a sovrin DID
type errors so that is probably the validation fix @swcurran and @dbluhm are mentioning.
However the behaviour of the issue seems to remain (mostly) the same as described in this issue.
For single and multi use OOB invitations I can accept them in the BC Wallet and the connection still appears to be finishing on the wallet side (though it does spin on the "making a connection screen), however on the Traction tenant I still see the connection only progressing to response
and messages sent from the wallet don't appear to go through.
So might be additional issue that needs investigating on the BC Wallet (or Bifold? or Credo?) side with sending a response back... not sure. Can follow up at a later time
So the validation error seen on ACA-Py is no longer there, but I will leave this ticket open until determining cause or raising with wallet/credo side.
Traction logs when scanning OOB invitation with BC Wallet
2024-03-07 23:58:06,971 aiohttp.access INFO 10.97.210.159 [07/Mar/2024:23:58:06 +0000] "GET /connections HTTP/1.0" 200 1816 "https://pr-1025-traction-tenant-ui-dev.apps.silver.devops.gov.bc.ca/" "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36"
2024-03-07 23:58:19,068 aries_cloudagent.protocols.routing.v1_0.manager INFO >>> fetching routing record for verkey: CGbKM9Nu88psngDFcuB41K7EviTWqMQuW3WMWMBvDXVM
2024-03-07 23:58:19,073 aries_cloudagent.protocols.routing.v1_0.manager INFO >>> FOUND routing record for verkey: CGbKM9Nu88psngDFcuB41K7EviTWqMQuW3WMWMBvDXVM
2024-03-07 23:58:19,078 aiohttp.access INFO 10.97.28.1 [07/Mar/2024:23:58:19 +0000] "POST / HTTP/1.1" 200 149 "-" "okhttp/4.9.2"
2024-03-07 23:58:19,082 aries_cloudagent.connections.base_manager WARNING No corresponding DID found for sender verkey: 3YuEyUAD4oMmpGiFVah5ytvPGwkGGJxz2jd9AcagVigF
2024-03-07 23:58:19,084 aries_cloudagent.connections.base_manager WARNING No corresponding DID found for recipient verkey: CGbKM9Nu88psngDFcuB41K7EviTWqMQuW3WMWMBvDXVM
2024-03-07 23:58:19,169 traction_plugins.traction_innkeeper.v1_0.endorser.endorser_connection_service INFO > connections_event_handler
2024-03-07 23:58:19,170 traction_plugins.traction_innkeeper.v1_0.endorser.endorser_connection_service INFO < connections_event_handler
2024-03-07 23:58:19,185 aries_cloudagent.multitenant.route_manager INFO Add route record for recipient H2cvtePmYrRbAncRP17yGuwd2T4X5nEM4Bw8w3ZYgzNn to wallet 30d67ebf-a775-4132-8995-e1b049addca8
2024-03-07 23:58:19,186 aries_cloudagent.protocols.routing.v1_0.manager INFO >>> creating routing record for verkey: H2cvtePmYrRbAncRP17yGuwd2T4X5nEM4Bw8w3ZYgzNn
2024-03-07 23:58:19,190 aries_cloudagent.protocols.routing.v1_0.manager INFO >>> CREATED routing record for verkey: H2cvtePmYrRbAncRP17yGuwd2T4X5nEM4Bw8w3ZYgzNn
2024-03-07 23:58:19,195 traction_plugins.traction_innkeeper.v1_0.endorser.endorser_connection_service INFO > connections_event_handler
2024-03-07 23:58:19,195 traction_plugins.traction_innkeeper.v1_0.endorser.endorser_connection_service INFO < connections_event_handler
2024-03-07 23:58:22,418 aiohttp.access INFO 10.97.210.159 [07/Mar/2024:23:58:22 +0000] "GET /status/ready HTTP/1.0" 200 172 "-" "kube-probe/1.26"
@ianco perhaps you can work on this with @loneil as you wrap up your “reuse” work. This is in the same area. Ideally, you can do a Zoom session together so @loneil can demonstrate the issue in an setup environment and @ianco can see what is going on.
I think https://github.com/hyperledger/aries-cloudagent-python/pull/2748 and AFJ/Credo version >= 0.5.0 will be required to complete an OOB exchange with ACA-Py from the BC Wallet
If we add support for emitting did:peer:1 to ACA-Py, we could get away with AFJ/Credo versions < 0.5.0
This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 5 days.
This issue was closed because it has been stalled for 5 days with no activity.
This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 5 days.
This issue was closed because it has been stalled for 5 days with no activity.
Reopening as we're on 0.12.x (and trying 1.0.0) and Credo 0.5.x and still seeing with BC Wallet. Will track this for investigations
Tested in https://github.com/bcgov/traction/pull/1279
Fix from https://github.com/hyperledger/aries-cloudagent-python/pull/3050 did the trick
Tracking the issues I've seen scanning invitation QR codes with the BC Wallet when creating an invitation through the Tenant UI. Not sure if these are a specific ACA-Py problem, or if it's an issue with the BC Wallet app? If existing issues I can close this one.
This is using Traction in the Dev environment with a Tenant that's on the bcovrin-test ledger.
Same type of issue with both Single and Multi use. Not setting anything in the optional settings.
Note we are able to connect Traction Tenant to (another) Traction Tenant through OOB successfully.
Creating an OOB single use from the TenantUI and scanning QR with Wallet
state: invitation
There is a problem in the callback in the Traction Tenant ACA-Py logs when scanning.
ValueError: Bad specification did:peer:1zQmR4vLev39FVRRWfoJQp9BjfEu1CuHnxo69qLqwPnE4jJJ does not correspond to a sovrin DID
Creating an OOB multi use from the TenantUI and scanning QR with Wallet
request
state