Sure they CAN be combined with multi-sig and oracles. But if this is not ALWAYS done (and it appears that it is likely to in fact be rare) then each SPV proof that does not so combine them, increases the incentive for a 51% attack (with miners not running OP_SIDECHAINPROOFVERIFY support, due to the ability to unlock SPV locked coins, and perhaps make some use of them).
If you have big enough user base then even 100% hash power attack cannot spend this BTC b/c it is not valid transaction.
1. you can ignore OP_SIDECHAINPROOFVERIFY
2. if some miner will add transaction what will spent BTC from address using OP_SIDECHAINPROOFVERIFY into block without valid "SideChain proof" then this block will be ignored b/c it will contain invalid tx.
Thanks, I've been intrigued on the details of that since the San Jose 2013BF conference.
This was something that Eli Ben-Sasson worked out
https://www.youtube.com/watch?v=YRcPReUpkcUThe TX are invalid for nodes running OP_SIDECHAINPROOFVERIFY, but if not...
This is an implementation of this SCIP (Succinct Computational Integrity and Privacy) so it has to be checked, yes?
It struck me as the opening of the window for deterministic computing.
Essentially verifiability of what all computers are doing all the time.
It sets up a powerful framework. Terrifying, important, and exciting.
I get why the core devs want to work on it. New math changes things.