Combining preconfirmations with based rollups for synchronous composability

5

Like for each L1 slot S, the N’th (certified and) sequenced L2 block with l1_slot = S is definitionally the slot-ending block. The sequencer can still prevent the synchronous composability window by publishing < N blocks but at the cost of degraded throughput / reputation / UX / etc.

Yeah I think doing this is reasonable. You can’t prevent the sequencer from stalling (I guess you can have an onchain mechanism that if no new block appeared on L1 for >10 slots, then it automatically enters based mode and any L2 preconfs since the previous based block are then invalidated, ie. equivalent consequences to a short reorg), but you can make it unattractive for them to do that.

2 Likes