Nicholas Clarke

Results 22 issues of Nicholas Clarke

Currently we have stabilityWindow and randomnessStabilisationWindow set to 3, giving us a 3-4-3 epoch structure. It was suggested by Peter Gazi that 4-2-4 would be a better parametrisation.

:microscope: research

Depends on https://github.com/input-output-hk/cardano-prelude/issues/115

:credit_card: technical-debt

# Test re-org - Make it obvious how to contribute additional testing - What should we do with the ConcreteCryptoTypes module? It's not parametrised everywhere anyway! Let's just drop it...

:credit_card: technical-debt

We use `Natural`/`Integer` in various places in the spec, most likely for consistency with the formal spec. But this is a little annoying for implementation, where we'd like to make...

:credit_card: technical-debt

`ShelleyGenesisStaking` results in internal manipulation of the initial ledger state, and is not intended for production use. It should be made optional, and we should avoid doing any manipulation in...

:credit_card: technical-debt

The exec spec currently makes use of lots of positional parameters. This was done to promote consistency with the spec, but can be confusing even in the formal spec (see...

:credit_card: technical-debt

:fountain_pen: documentation

As per https://github.com/input-output-hk/haskell.nix/blob/master/overlays/tools.nix#L38, a shell tool should accept a modules argument which gets passed to `cabalProject`. However, as far as I can tell, this argument is entirely ignored - it...

Given an executable qualified by a directory, `snack` complains about an invalid character in the derivation name: With this `package.yaml` section: ``` executables: ffexecutord: main: app/FFExecutorD.hs ``` ``` [nc@lorien:~/topos/tweag/funflow/funflow]$ snack...