Alter config filed name to devnet if it's not populated in file

[terencechain]

Today, Prysm ships with an embedded genesis state. When loading custom genesis state with config name unchanged (i.e., mainnet), prysm will by default uses embedded genesis state which conflicts with to be loaded genesis state. (i.e. EMBEDDED_GENESIS_STATE.HTR() != LOADED_GENESIS_STATE.HTR()). This is not ideal, and it's not quite right to assume that whenever config is unchanged, it's always going to use EMBEDDED_GENESIS_STATE. It breaks much of the assumption of starting a new chain from scratch. The assumption that uses EMBEDDED_GENESIS_STATE should be further restricted where deposit contract address and genesis fork version matching

EDIT:
Worked with @nisdas to find a more neutral way to alter mainnet field name in the event if it doesn't exist

credit to @parithosh for pointing this out

[nisdas]

Not sure if this is the right approach, if we want the user to be able to overwrite the embedded state then maybe we should have clearer logs indicating that it is indeed what we are doing. If users are providing a config with mainnet specified then it is reasonable to assume that mainnet is the config that they are using.

[terencechain]

Not sure if this is the right approach, if we want the user to be able to overwrite the embedded state then maybe we should have clearer logs indicating that it is indeed what we are doing. If users are providing a config with mainnet specified then it is reasonable to assume that mainnet is the config that they are using.

It's not users are providing config with mainnet specified, it's users providing config with nonespecified. In this case, it defaults back to mainnet config and loads mainnet genesis state