@viitalik-eth
Yes the core mechanism (PeerDAS + erasure coding) ensures robustness probabilistically, but it alone isn’t a magic guarantee. The harder part is making sure rational node operators have sufficient incentive to continue storing, serving, and not dropping chunks. We’re actively working on that factoring in penalties, rewards, redundancy, randomness, and dynamic protocol parameters. Over time, as the system matures, I expect the incentive layer to evolve in response to real-world stress and node behavior…