Change wallet loading code to access settings through the Chain interface instead of writing settings.json directly. This is for running wallet and node in separate processes, since multiprocess code wouldn’t easily work with different processes updating the same file.
This PR is part of the process separation project. The commit was first part of larger PR #10102.
Change wallet loading code to access settings through the Chain
interface instead of writing settings.json directly.
The following sections might be updated with supplementary metadata relevant to reviewers and maintainers.
Reviewers, this pull request conflicts with the following ones:
If you consider this pull request important, please also help to review the conflicting pull requests. Ideally, start with the one that should be merged first.
ACK 49ee2a0ad88e0e656234b769d806987784ff1e28 (jamesob/ackr/22217.1.ryanofsky.refactor_avoid_wallet_co)
Change seems straightforward and benign. Just want to confirm that there’s no need to update the two other calls to updateRwSetting in wallet/wallet.cpp (AddWalletSetting/RemoveWalletSetting).
ACK 49ee2a0 (
jamesob/ackr/22217.1.ryanofsky.refactor_avoid_wallet_co)
Thanks!
Change seems straightforward and benign. Just want to confirm that there’s no need to update the two other calls to
updateRwSettingin wallet/wallet.cpp (AddWalletSetting/RemoveWalletSetting).
That’s right. It’s safe for the wallet to use the Chain interface to update settings file, because the Chain interface methods are callable from the bitcoin-wallet process but execute in the bitcoin-node process. It is not safe for wallet code to use the gArgs methods to read and write json settings, because then the bitcoin-wallet process would be accessing the same settings as the bitcoin-node process and corrupting them or getting them out of sync between the two processes.
