From how Walrus is structured, slashing is expected to be progressive rather than binary. A single missed proof or brief outage should not erase a staker’s entire position. Instead, penalties are likely to scale with the severity, frequency, and duration of failures.
This approach matters because binary slashing favors large operators who can absorb risk more easily, which can quietly drive centralization. Progressive slashing allows smaller operators to recover from isolated faults while still discouraging persistent underperformance.
Parameter selection—such as penalty percentages, grace periods, and escalation thresholds—is ultimately a governance question. The goal is balance: strong enough incentives to protect availability, but not so harsh that only well-capitalized actors can participate safely.
From my perspective, this restraint is essential if Walrus wants decentralization to remain practical rather than theoretical.