On Wed, May 24, 2017 at 9:50 AM, Tier Nolan <tier.nolan@gmail.com> wrote:

> OP_BRIBE_VERIFY could then operate as follows
>
> <block height> <sidechain_id> <critical hash> OP_BRIBE_VERIFY
>
> This causes the script to fail if
>   <block height> does not match the block height, or
>   <critical hash> is not the hash for the sidechain with <sidechain_id>, or
>   there is no hash for that sidechain in the block's coinbase
>
>
I was thinking more on the process for these transactions.

I assume that the process is

- sidechain miner broadcasts transaction with OP_BRIBE output
- this transaction ends up in the memory pool of miners
- Miners add the transaction to their next block
- Miners add a transaction which spends the output to one of their own
addresses

I think you need an additional rule that OP_BRIBE checks fails unless the
output is locked 100 or more blocks.

The output script would end up something like

IF
   <block height> <chain_id> <critical hash> OP_BRIBE_VERIFY
ELSE
  <public key> OP_CHECKSIG
ENDIF

This output acts like "anyone can spend" for the one block height.
Otherwise, only the sidechain miner can spend the output.

This allows the sidechain miner to reclaim their coins if the transaction
ends up in a different block.

OP_BRIBE_VERIFY would have an additional rule

The script to fails if
  one or more of the transaction outputs start with something other than
the template
  <block height> does not match the block height, or
  <critical hash> is not the hash for the sidechain with <sidechain_id>, or
  there is no hash for that sidechain in the block's coinbase

The template is
  <100> OP_CHECKSEQUENCE_VERIFY