Open marc0olo opened 2 years ago
Compared with ETH, AEX2 can indeed be optimized, and ETH is more convenient to use
Thread-comment: https://forum.aeternity.com/t/easier-superhero-wallet-integration-in-js/9269/18?u=subhod-i
requestNode
featureRpcAepp
stamp. Route requests through wallet(Provide an example).Browser wallet, DeepLink wallet, and Hardware wallet
types.@kenodressel @thepiwo shall we use this issue for the further wallet-connection investigation?
sure, why not
I'm planning to do a minor improvement here: onSign
callback in AeSdkWallet
should be removed and signTransaction
method of Account used instead.
The reason is that in case we using Ledger HW the signing confirmation actually happens on the device side (while calling signTransaction
). In current architecture, we have either ask user to confirm signing twice (in onSign
and on Ledger side) or automatically allow the request in onSign
depending on the account type and then the rejection won't be handled properly.
In preparation for this, I've implemented Ledger support on sdk side: https://github.com/aeternity/aepp-sdk-js/pull/1685
Requirements (Result workshop at CryptoCastle, 13.09.2022)
TODOs
Historical Discussion / Content
this was discussed many times and raised by @nikita-fuchs a lot of times. not sure what implications we have here and if AEX-2 (which is "withdrawn") is the right approach here.
this definitely needs some discussion. here an old forum thread about this:
we also have some other open issues regarding wallet-connection, signing and AEX-2: