As i shared in my comment above I don’t think a split around multiprocess binaries is ideal. But i believe it is important to refocus the main project on the node parts of the software, and i just see such a split as the only realistic and socially acceptable way of achieving that. Do you have a better suggestion?
I agree that raw transaction parsing, PSBT updating, script decoding, signing, etc… should not be part of the node RPC since it does not required a node in the first place. But yes it’s tangential.
BDK does not assume that anymore. It now just indexes wallet transactions and has a wallet-spk-to-derivation-index index just like the Core wallet. It has a bitcoind RPC chain data source as well.
Me too. Unfortunately with the current resources it comes at non-trivial cost for all Bitcoin users, who are indirectly using Core, for the benefit of a handful of direct Core GUI users.
We have MiniWallet for this?
In a split future this could just be done using the Core wallet project which would release the equivalent of today’s bitcoind? Or just any other wallet that would bundle a bitcoin-node (Sparrow, Liana, …).