RPC: Return external_signer command in getwalletinfo

[kristapsk]

Alternative / follow-up to #24307. As was proposed by @luke-jr (#24307 (comment)):

Seems like this should probably be a string indicating the external signer program...

"external_signer" is changed from bool to optional string and is returned only if WALLET_FLAG_EXTERNAL_SIGNER is set. Will return empty string if wallet is configured to use external signer but Core is started without -signer.

[achow101]

This doesn't seem like it's really useful, and perhaps it can be misleading? The value will change depending on what -signer is set to, and it will always be the same for all wallets. Furthermore, because -signer is an explicit option, I would expect the user knows that information already. This could also be misleading and cause people to think that the value there is stored in the wallet itself rather than a startup option.

[w0xlt]

I think #24307 is a better approach.

[luke-jr]

This could also be misleading and cause people to think that the value there is stored in the wallet itself rather than a startup option.

IMO it should be the former eventually

[kristapsk]

Alternative approach would be to close this, keep the way of #24307 and later add another return value 'signer_program" to result object of getwalletinfo. Otherwise, if "external_signer" has to be changed at some point in the future, better to do it sooner, not introduce incompatible RPC change later.

[DrahtBot]

The following sections might be updated with supplementary metadata relevant to reviewers and maintainers.

Conflicts

Reviewers, this pull request conflicts with the following ones:

• #21928 (wallet: allow toggling external_signer flag by Sjors)

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.

[achow101]

This could also be misleading and cause people to think that the value there is stored in the wallet itself rather than a startup option.

IMO it should be the former eventually

Sure, but I think we should only add this output at that time rather than now.

[kristapsk]

Closing, agree this can be added later.