@vjudeu
> better to allow transaction joining.. to make fees more smoothly

I'm not familiar with RSK transaction joining. However, I don't think this addresses the issues Corey brought up - which is that the appropriate amount of security (ie miner revenue) isn't linked with any bitcoin market behavior. It sounds like you're suggesting something that could smooth out the fee market and potentially lower fees, however it doesn't sound like a mechanism that could be used to target a particular security level.

>  I think the market should adjust fees, and finding the right balance between on-chain and off-chain should be left to the users, just by providing them options like transaction joining

The market has an incentive to minimize fees. So I don't see how this would be sufficient if it ends up that miner revenue from fees is too low. 

@Erik
> I think perhaps you're  underestimating the degree to which utility can be added to the main chain to encourage fees.

It seems you've misinterpreted me to be saying that fees will be too low. I have no idea if fees will be too low or not. And you might be right that fees are likely to remain high enough. However, fees might also remain *too high* which itself could be a problem. As I noted above, since market forces all incentivize driving fees down, what do we do if that force drives it below a sufficient level? How will we know when that happens until we know what the sufficient level of security is (eg in terms of total miner revenue).

@Guiseppe
> If it only takes a few million dollars to attack BTC and make it completely unusable for one day

It would take much more than a few million to 51% attack bitcoin. Bitcoin's blockchain security is primarily based on the capital cost of buying up a massive amount of hashpower. The ongoing maintenance and electricity costs of mining are actually not very relevant to security because all those ongoing costs are directly offset by mining revenues. Its the upfront costs that serve as the primary barrier an attacker must surpass. Acquiring the mining hardware, finding places where energy costs are low enough, setting up the equipment, etc. To do that costs billions of dollars, not millions. 


On Thu, Jul 7, 2022 at 9:44 AM Giuseppe B via bitcoin-dev <bitcoin-dev@lists.linuxfoundation.org> wrote:
It's the first time I read about the security budget and it definitely sounds scary to me.
If it only takes a few million dollars to attack BTC and make it completely unusable for one day, I suppose it's only a matter of time before some hedge fund actually does it, using a short position to profit from the huge panic sell wave and global loss of confidence that would result from it. 
It seems even cheaper to do than the recent attack to Terra, unless I'm missing something. 


On Wed, Jul 6, 2022, 1:10 PM <vjudeu@gazeta.pl> wrote:
> If the only realistic (fair, efficient & proportionate) way to pay for Bitcoin's security was by having some inflation scheme that violated the 21 million cap, then agreeing to break the limit would probably be what makes sense, and in the economic interest of its users and holders.

So, Paul Sztorc was right again, there are three options: Enormous Block Size Increases, Violate 21M Coin Limit, or >50% Miner Fee-Revenues Come From Merged Mining: https://www.truthcoin.info/images/sb-trilemma.png. And I think using Merged Mining is the best option. More about that: https://www.truthcoin.info/blog/security-budget-ii-mm/

> Another option, if we were to decide we are over-secured in the short term, would be to soft-fork in a reduction in the current and near-future mining rewards, by somehow locking the coins in a contract that deprived the miner of the full reward, and then using that contract to pay the rewards out far in the future, should at some point we feel the security budget was insufficient.

Yes, that's also possible, RSK uses that. And making some kind of soft-fork for that is also possible, but I don't know if miners will agree to send some coinbase reward to "<futureBlockNumber> OP_CHECKLOCKTIMEVERIFY OP_DROP OP_TRUE".

On 2022-07-06 06:29:18 user Corey Haddad via bitcoin-dev <bitcoin-dev@lists.linuxfoundation.org> wrote:
>Bitcoin's finite supply is the main argument for people investing in it, the whole narrative around bitcoin is based on its finite supply. While it has its flaws and basically condemns bitcoin to be only used as a store >of value (and not as a currency), I don't think it's worth questioning it at this point. 
>
>Just my 2 sats. 
>
>Giuseppe. 


A finite supply alone is not enough to give something value, as it must also be useful in some way. In the case of Bitcoin, various forms of cryptographic security must all work - and work together - to make Bitcoin useful. If the only realistic (fair, efficient & proportionate) way to pay for Bitcoin's security was by having some inflation scheme that violated the 21 million cap, then agreeing to break the limit would probably be what makes sense, and in the economic interest of its users and holders.

There will always be competitive pressures with respect to efficiency, and both being over-secured and under-secured would be economically inefficient for a crypto currency, and thereby laving room for a more optimally-secured competitor to gain ground. Currently there is zero feedback in the Bitcoin system between what we might think is the optimum amount of security and what actually exists. There is also zero agreement on how much security would constitute such an optimum. Figuring out how much security is needed, or even better, figuring out a way to have a market mechanism to answer that question, will be an important project.

Another option, if we were to decide we are over-secured in the short term, would be to soft-fork in a reduction in the current and near-future mining rewards, by somehow locking the coins in a contract that deprived the miner of the full reward, and then using that contract to pay the rewards out far in the future, should at some point we feel the security budget was insufficient. Anthony Towns presented a form of this concept in greater detail at a Scaling Bitcoin conference some years ago. While this solution, if employed, would only work for some finite amount of time, it is possible that could give additional decades before the accumulated security budget was exhausted. 


Corey
_______________________________________________
bitcoin-dev mailing list
bitcoin-dev@lists.linuxfoundation.org
https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev