---
date: 2025-03-05
location: https://meet.google.com/vnz-chhb-qxz
recording:
---

## Agenda

- [x] HackMD team to collaboratively take notes and prepare agenda, example: https://hackmd.io/@cip-editors
- [x] Pi: Gummiworm
- [x] Hinson: DeltaDefi issues / experience report
- [ ] Tze Wai: Quick update on Mesh hydra tools
- [ ] Trym: Hydra product roadmap
- [ ] Noon: Hydra development update (etcd)

## Attendees

- Sebastian Nagel
- Perry Wasserbauer
- Pi Lanningham
- Jingles K
- Hinson Wong
- Noon
- Pedro Lucas
- Tudor Cotruta
- Trym Bruset
- Tze Wai Wu
- Oleksii Khodakivskyi
- Aaron Boyle
- George Flerovsky
- Franco Testagrossa

## Meeting notes

- Talked about using hackmd as note taking and this already works great as we can type all the same time!

Gummiworm:
- Motivated by explaining L1 and Hydra trade-offs
- L1 is very decentralized, but slow
- Hydra is very fast, and fully secure, but requires "small world"
- Key idea: split responsibilities of execution and custody
- Short term custody vs. long term custody
- People are fine with more centralized custody in the short term
- For example in exchange of fast user experience
- Multi-frequency Hydra
- at least two levels of confirmation
- Validation should be pluggable
- Big multi-sig at first
- stake-based or zk solutions thinkable too
- Being able to check the health of the system
- also to create back pressure
- Good behavior incentivized through slashing = optimistic protocol
- This makes a lot of sense on Cardano as SPOs could collateralize their pledge

DeltaDefi:
- Developing ongoing over last couple months
- Identified critical issues
- Working on Hydra in a way that making sure the software works before care about decentralization (aka Want to ensure no value locked if everyone is honest
- Would move towards more decentralized approach after proving the product is needed by community
- Brought up critical issue: https://github.com/cardano-scaling/hydra/issues/1773
- We should make sure its linked up with our latest idea how to solve it: https://github.com/cardano-scaling/hydra/issues/1858
- Sebastian pointed out that having non-finalizable head states is REALLY unsafe
- Maybe partial fanouts could help? https://github.com/cardano-scaling/hydra/issues/1468
- Hinson: We think incremental decommit can serve same purpose?
- Other two issues they noticed and will be closely monitoring:
- https://github.com/cardano-scaling/hydra/issues/1871
- https://github.com/cardano-scaling/hydra/issues/1757