PR #686 was merged after I had the opportunity to discuss the following scenario:
Scenario:
- Wallet 1.0 expects
invalid_grant for an invalid transaction code (before the PR was merged)
- Wallet 1.1 expects
invalid_tx_code for an invalid transaction code
- Issuer 1.0 returns
invalid_grant for an invalid transaction code (before the PR was merged)
- Issuer 1.1 returns
invalid_tx_code for an invalid transaction code
Assuming that mixed Wallet and Issuer versions need to interoperate, should we introduce additional considerations for handling invalid_tx_code? I do see interoperability issues in some scenarios, and I no longer recall why they were considered acceptable. I created the PR based on these assumptions but I feel we should at least document them.
Furthermore, should we revert the text in the 1.0 errata to the original wording, where invalid_grant was returned in case the transaction code was invalid?
Additionally, @paulbastian raised a point that we would also need an additional error code for here #686 (comment):
error code can be also used when correct tx_code was entered too many times. PR might follow.
@paulbastian @c2bo @jogu @fkj @javereec @Sakurann
PR #686 was merged after I had the opportunity to discuss the following scenario:
Scenario:
invalid_grantfor an invalid transaction code (before the PR was merged)invalid_tx_codefor an invalid transaction codeinvalid_grantfor an invalid transaction code (before the PR was merged)invalid_tx_codefor an invalid transaction codeAssuming that mixed Wallet and Issuer versions need to interoperate, should we introduce additional considerations for handling
invalid_tx_code? I do see interoperability issues in some scenarios, and I no longer recall why they were considered acceptable. I created the PR based on these assumptions but I feel we should at least document them.Furthermore, should we revert the text in the 1.0 errata to the original wording, where
invalid_grantwas returned in case the transaction code was invalid?Additionally, @paulbastian raised a point that we would also need an additional error code for here #686 (comment):
@paulbastian @c2bo @jogu @fkj @javereec @Sakurann