Shinobi’s Strawman is a weekly sequence the place our Technical Editor Shinobi challenges the Bitcoin neighborhood, aiming to fire up dialog round heated technical debates.
_________________________________________________________________
Right here is a component two of the experiment. Final week I printed a brief immediate difficult readers to reply with their very own protection or criticism of drivechains. The aim of this was to instigate challenges to my very own criticisms, questions, and even new criticisms I’ve not considered or thought of. Written type content material is mostly extra thorough and simpler to digest than real-time communication, as each events have time to take a seat and assume earlier than formulating a response versus needing to take action instantly. I feel this can assist to vary the tone of conversations round contentious subjects by attempting to facilitate them on this format.
In order that stated, time to undergo the responses to final week’s immediate.
Paul Sztorc
Paul Sztorc responded in lengthy type on Twitter, everything of which could be discovered here. For formatting readability in quote snippets, daring textual content is denoting which of my statements Paul is responding to.
> 1) Drivechains introduce a hodgepodge of recent variables into miners’ incentives … Drivechain is similar to RIOT’s use of “energy curtailment credit”. https://riotplatforms.com/bitcoin-mining It’s only a new method for miners to become profitable. Once I’m requested: “does drivechain have an effect on miner incentives?” I say “no”. I personally lived by way of the invention of: FPGAs/ASICs, heat-reuse, stranded natgas flaring, curtailment credit, and a complete lot else. Merged mining was invented by Satoshi in 2010, and is already in steady use — https://truthcoin.data/weblog/security-budget-ii-mm/#c-its-too-late–mm-is-already-widespread . Identical with the withdrawals — miners do loads of good issues, similar to MASF activate delicate forks or maintain peoples mistaken price cash ( https://x.com/satofishi/status/1701042302238724512?s=20 ), or rent Bitcoiners to shill Bitcoin. So, to somebody like me, getting revenues from merged mining, or overseeing 4 fully-automated withdrawals per sidechain per yr, does not even register as a change. It is simply enterprise as common.
Paul claims that energy curtailment agreements are equal to the centralizing pressures of drivechains. This can be a damaged comparability for just a few causes, first of which is the wild distinction when it comes to scale. One thing like working infrastructure for drivechains, or the proportional benefit of pool dimension in doing so, runs on economies of scale. The bigger an operation partaking in such a habits is the extra of a worldwide benefit it offers them. Energy curtailment however does not, it has diseconomies of scale. One mining operation partaking in energy curtailment on Texas’s grid has no affect in any respect on miners linked to a different grid having the ability to interact in related agreements. Mix this with mining actively getting used to increase renewable power manufacturing, which creates the necessity for these curtailment agreements, and your entire dynamic over time is assured to decentralize and grow to be an increasing number of open to different miners. Additionally, the declare that miners being put in absolute management of custodying different individuals’s funds and resolve which withdrawals to course of (by some means with out realizing the present balances of authentic customers) isn’t any change of their function is simply patently false.
> 2) Current Sidechains Have No Adoption Wait!? I believed sidechains had been going to vary miner incentives?? Not if they’ve “no adoption”. 😉 Anyway… RSK/Liquid are federated, and the federated mannequin is horrible. “federation vs PoW”, is actually the one distinction between Bitcoin (successful) and its failed predecessors. We will equally anticipate BIP300 to outcompete Federated. Moreover, they are not even in the identical league. Liquid doesn’t present us a web site (for instance) the place we will paste in (for instance) the zCash Altcoin supply code, and get out of {that a} zCash federated sidechain. As a substitute we’re caught with only one piece of closed supply junk that we can not modify. That misses your entire level of sidechains. Evaluating RSK/Liquid to Bip300 is evaluating two handwritten books to the printing press. Liquid was utterly closed supply till very just lately; nobody is aware of who the federation members are (regardless of the mannequin relying solely on their fame); the entire Liquid txn charges go solely to the company that created it. For some time (and nonetheless to today, for my part), Blockstream engineers might abscond with the funds if they really put 5 man-hours into it (see https://x.com/_prestwich/status/1277089486111817728?s=20 ). RSK aspires to be a drivechain — so I’ve their vote, a minimum of. They agree with me that they need to be a drivechain, not federated. Lastly, the truth that we have now did not construct issues that the end-user enjoys? That ought to solely spur us onward, to invent new issues. Not surrender quicker.
I do not know what to say right here…primarily each declare right here is fake. Liquid/Parts the platform has at all times been solely open supply and doable to switch, solely the code the federation members run to signal blocks and withdrawals was closed, however that’s now open supply. Paul pretending and attempting to indicate your entire mission was closed supply isn’t true. As nicely, the declare that “5 man hours” might steal the entire funds is solely false. The incident that he’s referring to was a bug (that has been patched) within the federation member code. All Liquid cash have a timelocked restoration path utilizing a 2-of-3 keyset within the occasion of catastrophic key loss by federation members that will end in all funds being misplaced. To ensure that these keys for use, the Federation should fail and stop transferring these UTXOs. That isn’t “5 man hours” of labor as Paul claims, it’s attacking a globally distributed set of HSMs which can be extremely sturdy to distant assaults and nearly definitely require bodily entry to compromise.
> 3) Drivechains Exacerbate The Dangers Of MEV > MEV is one thing that’s doable on Bitcoin already … however … Drivechains open the door to arbitrarily advanced types of MEV on sidechains, MEV = “miner aspect hustle”. In different phrases, if I supply Foundry $20 to shine my sneakers, then that’s MEV. If Slush Pool sells t-shirts on the aspect, then that’s MEV ( spoiler alert they already do: https://store.braiins.com/merchandise/braiins-polo-shirt ). Miner’s essential hustle is ordering transactions and blocks — anything they do, is a aspect hustle. Clearly we do not need the 2 hustles to battle! I addressed such “cross chain MEV” way back, in 2016, lengthy earlier than anybody had ever heard of shinobi (or MEV) ( https://youtube.com/watch?v=2OOKgTSrITs&listing=PLw8-6ARlyVciMH79ZyLOpImsMug3LgNc4&index=2 ). I designed Drivechain to have one thing referred to as “categorical management”, to *defeat* cross chain MEV …in contrast to for instance Blockstream’s simplicity which I imagine might exacerbate it (see Half 5 / code obfuscation ; or see http://truthcoin.data/weblog/contracts-oracles-sidechains/ http://truthcoin.data/weblog/drivechain-op-code/ http://truthcoin.data/weblog/wise-contracts/ for extra). Honestly although: MEV is a distraction. May a wise contract pay miners to reorg, or censor txns?? Sure. However a human, might additionally bribe a miner to do these issues. In the end it comes right down to: $ from txn charges, vs $ the attacker pays. Finest method to assist miners is to verify they’re wealthy — accumulating plenty of $ from the “essential hustle”. Ie plenty of merged mining.
I do not know what else to say besides that Paul continues to make absurd and excessive arguments right here. Promoting t-shirts requires new tools, new providers, new investments, whereas reusing your mining {hardware} does not. A miner selecting up a penny on the bottom doesn’t have any related affect to miner revenue or incentives, whereas somebody providing miners $10,000 per week to reuse their hashrate for a brand new goal does. Evaluating the 2 is absurd.
These are in reference to my reply https://twitter.com/Truthcoin/status/1699093434026406322 to his earlier article. I stand by every part in that reply! > …these simply shove the liquidity necessities onto one more celebration, assuming they are going to present large quantities of liquidity for nearly nothing in return Each halves of this are fallacious. First, on the L1 aspect of the commerce, nothing is locked up — EVERY coin on L1, is already “offering liquidity” (on this context). Second, they definitely do not get nothing! They cost a price. The mannequin can be: “shopping for 1 sidechain coin, for 0.99 L1 cash” (for instance). > don’t assume it is a foregone conclusion that sufficient liquidity to cowl the “resolution to the safety finances downside”
I feel Paul right here is oversimplifying what’s going on, and ignoring the dynamics of arbitrage, which is what is occurring right here. Sure, in an excellent situation, all mainchain cash can be found to swap for sidechains, however in actuality that’s not the case. That assumes everybody thinks drivechains are equivalently safe to the mainchain. In actuality, there’s a safety and danger distinction, and folks partaking on this arbitrage are bearing that danger on behalf of individuals they swap with. Most Bitcoiners usually are not taking their bitcoin and arbitrage buying and selling for yield with it, they simply maintain it. That will not magically change due to drivechains, and finally the individuals doing this arbitrage have to get the cash they’ve swapped into drivechains again out to the mainchain to shut the arbitrage loop. This merely shifts that bottleneck immediately from sidechain block constructors to arbitrage merchants. Additionally on the finish of the day, this provides one other reduce another person is taking from the price sharing, and is a margin that miners can seize by operating a sidechain node themselves.
idBrain
Anon idBrain on X (Twitter) posted the query what would I do if drivechains had been activated. Nicely, in most conditions nothing. A URSF (Person Resisted Delicate Fork) attempting to go up towards your entire ecosystem can be largely futile, i.e. if most customers, companies, and miners all supported activating the proposal. If solely miners activated it, with no customers or companies price mentioning implementing it, it is perhaps price it to constantly suggest withdrawal transactions, looting the sidechain and paying all of it out to miners. If 51%+ of miners defected from implementing the foundations all drivechains may very well be looted with no time delay in a single block. If it did efficiently activate with huge help although, I might most likely stop taking a look at Bitcoin as one thing that might realistically stand as much as state and alter the dynamics of cash and state. It might be merely a fiat denominated funding to me at that time on the street to state seize.
Mister Ticot
Mister Ticot despatched in an e mail a query:
You talked about sidechains arn’t getting used and are solely federated. What about Stacks? Does not it qualify as a permissionless side-chain with some degree of success?
I might not qualify or describe Stacks as a sidechain in any respect. I might name it a para-chain, or a parasite chain. Stacks is an unbiased community with a local base token completely different from Bitcoin, and as such I don’t qualify it as a sidechain. It interacts with Bitcoin in the same method, and by that advantage can affect Bitcoin miner incentives, however it isn’t constructed on a basis of BTC because the core native asset, which I feel is the principle requirement for a secondary blockchain to be thought of a sidechain.
Micah Warren
Micah Warren wrote in an e mail: Responding to your name to fire up technical dialog.
Responding to your name to fire up technical dialog.
My understanding is that the massive unavoidable havoc-wreaking downside with blind merged mining is that it is trivial to acquire as many blocks as you would like just by outbidding different ‘miners’. It rapidly degenerates right into a bluffing/signaling recreation. It additionally create conditions the place you may create large MEV alternatives by committing to longer reorgs, along with quick time period performs like fee-sniping. In proof of labor, if somebody tries to carry out an extended reorg, the sincere miners (supplied there’s 51%,) can simply default to the identical factor they at all times do. Nonetheless in BMM, when you decide to successful the public sale to hold out your shenanigan, there isn’t any default mode that sincere miners can retreat too. All unhealthy stuff. In my view, this makes BMM probably not a severe consensus mechanism.
HOWEVER, it most likely could be fixed- you simply should barely assume outdoors of the PoW field.
This is the factor, as a result of the map from SC blocks to L1 blocks is injective, we receive a linear, sequential, whole ordering of all candidate sidechain blocks. So actually, we’re 99% of the best way there so far as consensus goes – we have narrowed it down from trillions of doable blocks to a small discrete handful of candidate blocks and these blocks include a transparent whole ordering. The one factor fallacious with taking the primary block at peak N to be the canonical one is that such a block at peak N may not be legitimate. So all you want is a straightforward mechanism to find out, inside a brief time period, whether or not the block at peak N is right or whether or not it ought to be discarded. Clearly invalid blocks will finally be discarded, the one query is the best way to implement a time restrict so that somebody cannot maliciously withhold a block for a very long time so as to jam up consensus.
This does not seem to be a tough downside. One resolution: You might merely declare a jury of community-trusted sidechain nodes, say 5 of 9, who would wait 20 seconds after the block is mined, and if they will validate the underlying block, they are saying it is good, it is now within the canon. If they cannot see the block or cannot validate it, they declare it invalid.
Now the 20 seconds is bigoted, the jurors are simply calling balls and strikes, there does not must be an accurate reply – the one factor is that 21 seconds after the final L1 block has been mined, sidechain miners now know for positive whether or not to mine a brand new block or on prime of the outdated one.
Drawback solved. The one downside (laser-eyed maxis would possibly wish to ear muff for this), it’s important to depend on one thing apart from proof of labor to resolve uncommon consensus disputes. After all, such disputes would nearly by no means occur, as a result of the one purpose they might occur is that if an adversary was attempting to create a schism level, and by breaking the tie immediately, you’re obviating the schism level.
After all what occurs on a sidechain is the sidechain’s enterprise – but when I might argue that the very best design of a sidechain would at all times contain some reorg safety, then all of the considerations about chaotic reorging forcing the L1 miners to enter the sport are not legitimate.
In response to this statement, I might say a distinct potential resolution that’s superior can be a Zero Information Proof of correctness for commitments to new sidechain blocks. Nonetheless, I feel fixing this difficulty undermines one of many core targets of drivechains structure: to not introduce new causes or incentives for miners to reorg the mainchain to perform a reorg on a sidechain. Micah’s proposal for federating validity testifying to sidechain blocks would create the identical incentive, however moreover finally backstop your entire belief mannequin of the sidechain with a federation. I.e. nothing can be thought of legitimate with out the attestation of these chosen arbitrators. This defeats the aim of drivechains design, which is to have miners fill the function as the final word backstop within the belief mannequin.
Alright, so that’s it for this week’s Strawman. Subsequent week I’ll attempt to be extra triggering.
