Slashing

[AshwinSekar]

This discussion page serves as an avenue and roadmap to discuss implementing slashing in the solana protocol.

Current proposals

#180 Leader schedule migration

• Populates leader schedule via vote accounts rather than node id, prerequesite for SIMD-0212

#204 Slashable event verification

• This provides an on chain mechanism to report and record violations
• Currently only supports duplicate block violations, future SIMDs will add voting violations

#212 Slashing

• Implementation of slashing via burning delegations and rewards
• Requires governance vote

Accepted proposals

TODO

• SIMD-0180 out for review SIMD-0180: Leader Schedule Migration #180

• SIMD-0180 is accepted

• SIMD-0180 implementation started

• SIMD-0180 implementation complete

• SIMD-0204 out for review SIMD-0204: Slashable event verification #204

• SIMD-0204 accepted

• SIMD-0204 implementation started https://github.com/solana-labs/solana-program-library/tree/master/slashing

• SIMD-0204 implementation complete

• SIMD-0212 out for review SIMD-0212: Slashing #212

• SIMD-0212 accepted

• SIMD-0212 governance vote initiated

• SIMD-0212 governance vote completed

• SIMD-0212 implementation started

• SIMD-0212 implementation complete

[Andrew-DCA]

Hey there,

I've had a look through the slashing proposal and wanted to share my thoughts.

The team's done a solid job breaking this down into three main parts:

First up, there's the Leader Schedule Migration (#180). It's pretty straightforward - moving from node IDs to vote accounts. Makes sense as a starting point and sets everything up nicely for what's coming next.

The verification system (#204) looks good. Love that they're starting with duplicate block violations before tackling the more complex stuff. The fact that there's already some code in the SPL repo shows they're making real progress.

The actual slashing mechanism (#212) ties it all together. Good to see they're planning to run it through governance - gives everyone a chance to weigh in.

A few things I reckon would be worth considering:

• Could use more detail on how the penalties will be calculated
• Would be good to see how they'll handle any disputes that come up
• Some concrete testing criteria

Overall though, it's a solid proposal. The step-by-step approach makes sense, and it's good to see they're not trying to rush everything in at once. The dependencies between each phase are clear, and the governance involvement keeps everything above board.

I'd say it's worth moving forward with. The team seems to have their heads around the technical challenges, and the structure looks sound.

Keen to hear what others think about it.