fix: block parsing logic for public credentials#697
Conversation
|
I'd have to take more time to give qualified feedback, but looks reasonable on first glance. |
|
@rflechtner I understand and I agree it would make things easier, but remember that the credential is user-provided input. Yes, it is bound in the maximum size it can have, but the general feeling also on the forums is that events should be kept small and only used as index to something else in the state (e.g., https://substrate.stackexchange.com/questions/4296/large-event-payloads-unsafe). |
…c-credentials-block-parsing-fix
…c-credentials-block-parsing-fix
10c4a93 to
63e2c79
Compare
63e2c79 to
b756f70
Compare
|
@rflechtner since this is blocking BTE, I've decided to go on and merge it after getting a review from Tom. Anything else you wanted to discuss about this, we can do once we're both back, and address any further open points. |
Previous logic was assuming that the only extrinsic that could create a public credential was a
submit_did_callextrinsic. This is not true, asbatch,batchAll, andforceBatchcan also create a public credential.This is now taken into account in this PR, and an integration test that covers the new case has been added.
One problematic block, raised by BTE, was https://kilt-testnet.subscan.io/block/2128207.