-chain param
#23068
This change adds some aliases to the -chain param. -chain=signet
is accepted, and this changes it into accepting -chain=testnet too.
Stay a private API, no documentation is changed. Some tests could be added, or not, depending on the software’s code coverage policy.
This change adds some aliases to the `-chain` param. `-chain=signet`
is accepted, and this changes it into accepting `-chain=testnet` too.
Stay a private API, no documentation is changed. Some tests could be
added, or not, depending on the software's code coverage policy.
-testnet -regtest etc). If you do type it a lot and get it wrong every time, why not define an alias at the shell level.
I think it’s best to have one, unambiguous, name per chain.
Yes.
Consistently. main, test and sig; or mainnet, testnet and signet.
I would also imagine this option is mostly used by scripts, not by users (which tend to type -testnet -regtest etc).
Yea, I just thought of iterating prettier.
If you do type it a lot and get it wrong every time, why not define an alias at the shell level.
This made me realize, using shell variables I can almost do what I want. Good enough. Thank you! ❤️
Consistently. main, test and sig; or mainnet, testnet and signet.
I agree it’s not really pretty. Some of the identifiers come from long-defunct BIP70 (main and test), the rest were added ad-hoc. I think if you present it as a plan to deprecate the old names eventually and go to, say {mainnet, testnet[3], regtest, signet} it might be more acceptable than ‘just adding aliases’. I don’t know.
I think if you present it as a plan to deprecate the old names eventually and go to, say
{mainnet, testnet, regtest, signet}it might be more acceptable than ‘just adding aliases’.
Change: Replaces ‘main’ and ’test’ ‘-chain’ param values with ‘mainnet’ and ’testnet’ aiming increased consistency with the recently added ‘signet’.
Introduce ‘-chain’ param values ‘mainnet’ and ’testnet’. They are equivalent to ‘main’ and ’test’ before this change.
Deprecate ‘-chain’ param values ‘main’ and ’test’. Using them makes the software log and output a warning. Add a note (comment) asking for removing the deprecated values no later than five years after being deprecated (unless the deprecation is reversed), the years are written in the comment.
This is small enough it doesn’t require a BIP, but this description must be sent to the mailing list after the PR is not pending requested changes, long before actually merging it.
If this sounds airy, it is, I’m missing a lot of BIP reading, I haven’t even ever used regtest, I’m not familiar with the source code.
-chain.
1010@@ -1011,7 +1011,18 @@ std::string ArgsManager::GetChainName() const
1011 if (fTestNet)
1012 return CBaseChainParams::TESTNET;
1013
1014- return GetArg("-chain", CBaseChainParams::MAIN);
1015+ auto chain_name = GetArg("-chain", CBaseChainParams::MAIN);
1016+ std::string chain_name_post_alias;
1017+ if (chain_name == CBaseChainParams::MAINNET_ALIAS) {
I never really upgraded to C++11.
It’ll take me a while to figure this out.
laanwj commented: It wouldn’t need a BIP, nor mailing list discussion.
I’ll see about writing the code and submitting another PR.
The following sections might be updated with supplementary metadata relevant to reviewers and maintainers.
Reviewers, this pull request conflicts with the following ones:
main and test -chain params by katesalazar)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.
