Support Vasil features (v2 scripts, inline datums and reference inputs)

[bladyjoker]

TODO

• Migrate Plutarch code to use Plutarch.Api.V2
  • STATUS: DONE
  • Update Plutus design with $AA, $AUTH, $CERT and $FS tokens #32
• Migrate shared types to use PlutusLedgerApi.V2
  • STATUS: DONE
  • Update Plutus design with $AA, $AUTH, $CERT and $FS tokens #32
• Migrate BPI/Plutip PAB Contracts to use V2 scripts
  • STATUS: DONE
  • Update Plutus design with $AA, $AUTH, $CERT and $FS tokens #32
• FsMp Plutus script uses reference inputs
  • STATUS: DONE
  • (Re #41) Audit review #42
• Enable the use of v2 scripts in bot-plutus-interface
  • STATUS: DONE
  • [DO NOT MERGE] Enabling V2 scripts bot-plutus-interface#151
    • Works, but we need to expose the option to choose between v1/v2 in the config
    • Add vasil features bot-plutus-interface#157
• Enable the use of inline datums and reference inputs in bot-plutus-interface
  • STATUS: DONE
  • BLOCKERS:
    • plutus-contract and plutus-apps need to be extended with new contraints/lookups
    • Inline datum to Constraints API proposal PLT-682 ADR for the constraint interface change following supporting inline datums IntersectMBO/plutus-apps#660
    • mustReferenceOutput PLT-670: MustReferenceOutput IntersectMBO/plutus-apps#661
    • Add vasil features bot-plutus-interface#157

References

• Reference inputs in the transaction https://cips.cardano.org/cips/cip31
• Inline datums in the transaction https://cips.cardano.org/cips/cip32
• Add vasil features bot-plutus-interface#157

[GeorgeFlerovsky]

RE plutus-apps (and I think plutus-contract), you only need these dependencies for plutip (via bpi), right?
If so, then we are looking at removing these dependencies from BPI with our own alternatives. @MangoIV is the main person currently working on this effort.

Something to consider: CTL currently has Babbage support and plutip (non-BPI) support, so it may be less blocked than the alternatives.

[bladyjoker]

RE plutus-apps (and I think plutus-contract), you only need these dependencies for plutip (via bpi), right? If so, then we are looking at removing these dependencies from BPI with our own alternatives. @MangoIV is the main person currently working on this effort.

Correct yes! Essentially we need those new constraints to referencing outputs and adding inline datums. Apparently it's not so difficult to add so might be that I can get to working on that fork. I'll sync with @MangoIV.

Something to consider: CTL currently has Babbage support and plutip (non-BPI) support, so it may be less blocked than the alternatives.

I hope we're not hot swapping the PABs just yet. But we definitely need to assess the timelines here. Thanks @GeorgeFlerovsky for providing the bigger picture on the state of things, super valuable.

[GeorgeFlerovsky]

Haha, definitely let's avoid swapping pabs if we can. With the architecture you described in the meeting, I think it's eminently possible to do so, as long as bpi progresses quickly on de-plutus-apping