When a L2 reorg occurs the batcher should close out the pending channel
and submit everything that it has before clearing the channel builder.
The reason for this is that even though it may be submitting incorrect
data it should not leave dangling channels. After the pending state is
drained it can begin tracking the new L2 chain.

op-node: Remove beta-1 config

--network=beta-1 is no longer supported as the network has been shut down.

op-program: Embed goerli L2 chain config

op-program: Bootstrap program via local oracle

op-batcher,proposer: Default previously required flags

fix(op-chain-ops): Compute local output root with op-node exported function

[op-batcher] Add support for multiple batcher transactions per L1 block

proxyd: sliding window

 - Move pending metrics to txmgr package (and switch to int64)
 - Timeout context in tests
 - require.Duration
 - Fix comment docs