Remove era arguments to cardano-testnet
#6314
Conversation
nbacquey
force-pushed
the
nb/clean_cli
branch
5 times, most recently
from
671b114 to
9725a00
Compare
| let genesis = Api.alonzoGenesisDefaults (toCardanoEra sbe) | ||
| defaultAlonzoGenesis :: Either AlonzoGenesisError AlonzoGenesis | ||
| defaultAlonzoGenesis = do | ||
| let genesis = Api.alonzoGenesisDefaults ConwayEra |
There was a problem hiding this comment.
This is fine for now but we should parameterize everything on Era era instead. Era era tracks only the latest eras.
There was a problem hiding this comment.
Saved in #6317 to keep it in mind
There was a problem hiding this comment.
This is fine for now
@Jimbo4350 I'm not sure that it is. In order to reintroduce parametrisation on Era era again, we'd have to revert this PR basically. Is it possible to do this now and keep the parameterisation?
nbacquey
force-pushed
the
nb/remove_broken_link
branch
from
20c8407 to
46cbb65
Compare
|
Needs an update of golden files: https://github.com/IntersectMBO/cardano-node/pull/6314/checks?check_run_id=48999840783 |
| @@ -197,13 +192,13 @@ createSPOGenesisAndFiles | |||
| H.note_ $ "Number of stake delegators: " <> show numStakeDelegators | |||
| H.note_ $ "Number of seeded UTxO keys: " <> show numSeededUTxOKeys | |||
|
|
|||
| let era = toCardanoEra sbe | |||
| let era = toCardanoEra ConwayEra | |||
There was a problem hiding this comment.
I think this place may be hard to find, where the era is hardcoded. Also It will be useful to still be able to override this value, since we're starting work on the next era now.
There was a problem hiding this comment.
I see two problems here:
- Conway era is hardcoded in multiple places (as a protocol version too) (also in comments), it should be hardcoded in at most one place if we want to move forward with this change.
- We're losing parametrisation on era. This is especially useful when working two eras at the same time: when one era is on mainnet and the other one is in the works. To restore the parametrisation we'd have to basically revert most of this PR. Can we keep the parametrisation, but just use a default era and not expose it to the user for example?
My hard requirement is nr 1, but let's discuss nr 2.
Starting a testnet in an era different than Conway has not been possible for a while, since the options to do so have been removed from `cardano-cli`. This commit therefore removes those options from `cardano-testnet` as well. This commit also removes leftover golden files.
|
@carbolymer your comments have been addressed. Basically I only changed the parser, and added a single hardcoded value in |
I don't understand where that's coming from: I did regenerate the golden files, and that one exists: https://github.com/IntersectMBO/cardano-node/pull/6314/files#diff-bcd54a52a6f758e10284e243ba0830249a2d15898ce60d606ceead57c74941e3 |
@nbacquey Ah right, it's green in GHA. We have hardcoded paths in nix code, so you have to update those as well: |
Starting a testnet in an era different than Conway has not been possible for a while, since the options to do so have been removed from
cardano-cli. This PR therefore removes those options fromcardano-testnetas well, along wih their respective code paths.This PR also removes leftover golden files.
Checklist
CHANGELOG.mdfor affected packagehlint. See.github/workflows/check-hlint.ymlto get thehlintversionstylish-haskell. See.github/workflows/stylish-haskell.ymlto get thestylish-haskellversionghc-9.6andghc-9.12