2019-02-20 Meeting notes: validation and Casper

Date

Feb 20, 2019

Participants

  • @Kelly Foster

  • @Lucius Meredith

  • @Dominik Zajkowski

  • @Former user (Deleted)

  • @Chris Boscolo

  • @Artur Gajowy

  • @Former user (Deleted)

  • @Philipp Strauch

Goals

We will use this time to resolve questions and open issues in validator incentives and Casper.

Discussion topics

Item

Notes

Item

Notes

Resources

Lazy validator hedge (challenge protocol)

  • DECISION We will implement a lazy validator challenge protocol

    • Request to table to discuss the design at another time

  • Estimated a week’s worth of work

  • Discussion of liveness faults challenge

    • Idea 24 hours to respond

Paragraph 5 : “Transactions signed to a particular validator”

  • This is not desirable

    • Risk of pushing out other validators “popularity contest”

  • Ideal is a gossip network so that it doesn’t matter who you deploy to. There is no special relationship between client and validator

  • There is not a gossip network for Mercury. We want to mitigate the risk of preventing being able to implement it later

  • DECISIONS

    • remove “transactions are signed to a particular validator”

    • Go back to a stake-weighted proposal for distributing fees

Paragraph 6: “all validators are taxed”

  • DECISION remove tax no longer needed

Discussion of attack related to asymmetry of distribution of fees

  • Concern is over the lifetime of Mercury will we see a serious threat to the network?

    • Risk of centralizing block proposing

    • Creating an edge for some validators

Next steps

  • We will meet again to continue discussion

Action items

@Former user (Deleted) delete paragraph 6
@Former user (Deleted) Edit paragraph 5 to remove stipulation for signing transactions to a particular validator
@Former user (Deleted) Edit paragraph 5 to revert to stake-weighted distribution of fees

Decisions

  1. Convert values to parameters