pull down to refresh
Thanks! And yes, more wacky frankenminers are in the queue. :)
We agree on the 51% case. When a censor refuses to build on someone's blocks, fees and defectors will hopefully be the answer. No home miner will have enough hash to secure their own inclusion.
My worry is the more banal attack that doesn't even need 51%. Foundry and the AntPool group recieve a compliance mandate and simply leave certain payments out. Fees don’t fix that. First, the compliance is not for sale. No amount of fees covers a prison sentence or the company being shut down. Then second, the actual fee never reaches the ones doing the hashing. With an FPPS pool, the operator builds the template, so the mouth-wating fee goes to the party who's already said no at that point.
Fees are the signal, and constructing one's own template is how a miner can respond. Fees may determine who gets the uncensored blockspace, but its the miners who are building their own templates creating it. If all six of the biggest dogs comply, that’s like 6-7 blocks a day, room for only 5% of today’s transactions. Small guys already up and running is what makes your fee defense work without waiting months for the relocation or rebuilding of new hashrate.
No amount of fees covers a prison sentence or the company being shut down
...
relocation or rebuilding of new hashrate.
There is a difference between a government demanding compliance from a pool and demanding compliance from all the hashers pointed at the pool.
In the case of Foundry, most of their hashrate seems to be coming from US based hashers, but even Foundry says they have non-US hashrate.
In a censorship regime, what stops the non-US hashers from going to a pool that will go after higher fees?
A pool like Antpool seems like it may have hashrate from a much wider range of jurisdictions. If China demands compliance with censorship, what stops Antpool hashers from jumping to new pools? For instance, I believe Coolbit mines with Antpool but has mining operations in the US.
constructing one's own template is how a miner can respond
Isn't it the case that a miner can also respond by switching pools?
For sure, and switching pools is a big response. But switching pools doesn't mean switching template builders. AntPool’s group includes pools a lot of us like, Luxor and Braiins for example, so lots of miners who “left” for them are still hashing AntPool’s templates (unless they’re using SV2 on Braiins with job declaration). That’s why I put the tag-vs-builder flipper on the site. The switch that actually changes who decides the payments is to a template you or someone outside the AntPool mandate builds.
Will industrial hashers make that switch? I don't know any that will. Marathon mined OFAC-compliant blocks in 2021 without even being asked to do so. I know they stopped because of public outcry, but that was for voluntary censorship. I don't know if they'd be so persuaded under strong coercion. At Bitcoin 2024 an IREN rep told me unapologetically that their hash was split between Foundry and AntPool and said “I don’t know” when I questioned him about a 51% event. US industrial miners are as much within OFAC’s reach as the pool is. Meanwhile fees are currently half a percent of what blocks pay out, so a few censored payments are a rounding error to an FPPS miner’s income, against the risk of the full force of the federal government.
Non-US hashers exiting to pools that build their own templates is what I’m rooting for. The benefit of home miners is they’re already there when something goes down, and their own payments are on the line. Again, maybe it's not someone's exact payment that they're going to mine, but it's that they're contributing hash that's in their favor.
The site is really fun! Great job. (But even more, I'm loving your adventures in making various bitcoin waste heat devices -- these are truly excellent.)
You bring up an interesting point for me: I don't think controlling template construction is part of the bitcoin threat model.
I am not yet convinced that control of the template is really that important.
I imagine an ultimate censor pool: >50% of hashrate mines blocks that are empty (except for their coinbase) and refuses to mine on top of anyone else's blocks. I think the Bitcoin response to such a situation is that users will have to increase the fees they offer miners for inclusion in transactions until hashrate defects from the censor.
It is highly unlikely any individual will be able to deploy enough hashrate to guarantee the inclusion of their own transactions. Rather, they hope to entice a large enough percentage of miners to include their transactions by offering juicy fees.
If the argument is that higher fees will not entice miners away from the censoring pool because such miners are ideologically motivated (ie owned by governments that want to censor), then, yes, this is a 51% attack, and it is a known vulnerability of bitcoin.
But, as far as I understood, the argument for why Bitcoin mining actually has a chance to remain decentralized is because energy is widely distributed around the world and so when bitcoin holders offer more btc in fees, there is a chance that more miners will come online (or switch pools because they are not under the jurisdiction of the censor).
I think it is a good thing that miners can build their own templates, but I don't think that is where the buck stops with censorship resistance. It seems to me that the fees users offer are the only real defense we have against a 51% attack.