Skip to content

Data Availability Sampling -- Fork Choice

Notice: This document is a work-in-progress for researchers and implementers.

Table of contents

Introduction

This document is the beacon chain fork choice spec for Data Availability Sampling. The only change that we add from phase 0 is that we add a concept of "data dependencies"; a block is only eligible for consideration in the fork choice after a data availability test has been successfully completed for all dependencies. The "root" of a shard block for data dependency purposes is considered to be a DataCommitment object, which is a pair of a Kate commitment and a length.

Dependency calculation

1
2
3
4
5
6
7
8
9
def get_new_dependencies(state: BeaconState) -> Set[DataCommitment]:
    return set(
        # Already confirmed during this epoch
        [c.commitment for c in state.current_epoch_pending_headers if c.confirmed] +
        # Already confirmed during previous epoch
        [c.commitment for c in state.previous_epoch_pending_headers if c.confirmed] +
        # Confirmed in the epoch before the previous
        [c for c in shard for shard in state.grandparent_epoch_confirmed_commitments if c != DataCommitment()]
    )

1
2
3
4
5
6
7
def get_all_dependencies(store: Store, block: BeaconBlock) -> Set[DataCommitment]:
    if compute_epoch_at_slot(block.slot) < SHARDING_FORK_EPOCH:
        return set()
    else:
        latest = get_new_dependencies(store.block_states[hash_tree_root(block)])
        older = get_all_dependencies(store, store.blocks[block.parent_root])
        return latest.union(older)