I saw Ruben Somsen discussed it last year, and more people talking about it recently.
Is it possible to implement rollup on BTC without hard fork?
ZK-Rollups cannot be implemented in without a hardfork since Miners would have to verify a zero-knowledge proof and miners do not do this in Bitcoin. Bitcoin does have Hash-preimage verification since the very beginning - one could imagine a distributed hash & hash-preimage generation but there is no such algorithm to my knowledge. If there was it would require new and complicated cryptography or a trusted dealer - which both isn't in the spirit of Bitcoin.
Optimistic Rollups are not possible on Bitcoin since they require to undo a transaction in case any party challenges the result. They would also require miners to verify the rollup when someone challenges the result which miners do not do in Bitcoin.
I don't think any of this is a good idea. Rollups always end up with so much chaos that it isn't in the spirit of Bitcoin. And if a construction requires a trusted dealers or a trusted group of participants - why bother with it in the first place - just do it in a SQL database and write the result on the Blockchain.
reply
Rollups always end up with so much chaos that it isn't in the spirit of Bitcoin
Can you elaborate on that?
just do it in a SQL database
A SQL database is much more centralized and permissioned than a rollup I'm sure.
reply
Not OP but LN is just a set of SQL databases. Every node maintains their own channels.db and uses cryptographic proofs to maintain integrity with their peers.
reply
Yes, but that is an entirely different use case for a SQL database than what we are discussing here.
reply
Just do it in an SQL database and write the result on the Blockchain
Literally describing LN
reply
LN is a network of SQL databases that have to agree on state, yes.
reply
As far as I understand rollups are scams. The optimistic rollup arbitrum had to throw towel in the ring a few weeks ago when fees became higher than on ethereum. They stopped the event they were doing. Arbitrum is a PoA rollup so there is no consensus mechanism. Arbitrum gets all the fees and decide which transactions to include or not. Despite this centralization they still ended up with fees higher than base layer. They are working on optimizations but that's it.
Zkrollups can perhaps do 5x more transactions than optimistic rollups before the ceiling is hit. But it's next to nothing when you want to do everything in a blockchain manner and want millions of users but i could be wrong.
reply
"SOLVE SCALING WITHOUT COMPROMISE" this is arbitrums motto. Considering its a PoA architecture and they stopped the odyssey event a few weeks ago that motto seems like a lie. Sad.
reply
Yeah I'm not sold on optimistic rollup after the Arbitrum fee incident. However some people I follow on Twitter seems excited with Starkware (and they are working with HRF) so I want to verify with other Bitcoiners about Starkware's implementation as well.
reply
I don't know a lot about rollups, but seems like offloading computationally-intensive work off the blockchain is pretty antithetical to Bitcoin.
reply
Offloading work from the base layer seems consistent with choices the network has made in the past.
reply
Well, yeah, I guess that's true. Depends on how it's pulled off - for some reason I was thinking this was suggesting making mining easier, which would be antithetical, but utilizing an upper layer is kind of what LN is doing, right?
reply
Yep. LN exists precisely to take load off the base layer.
reply
Not without a valid reason, that's for damn sure
reply
I don't pretend to know all of the details, but I think it would require a hardfork.
I think zkrollups would be a fantastic way to increase demand for blockspace and steal marketshare from Ethereum.
Like it or not, Ethereum's shitcoin casino is generating a ton of demand for blockspace. And there are some legitimate things on ETH, too (like the ability to borrow against your (wrapped, unfortunately) BTC and use stablecoins).
reply