Reminder that a softfork, even a miner-activated softfork (MASF), is always user-enforced. Miners attempting to enforce a rule change on their own is what's called a 51% attack. It's impossible to have a covert softfork.

Replies (39)

techfeudalist 's avatar
techfeudalist 2 years ago
Please help me understand the BIP 300 risks better. All I’m hearing at the moment is that people don’t want sidechains to use BTC as a native currency. Pros: + gives miners more revenue + allows people to experiment with other functionality using their bitcoin Cons: … unclear How does this negatively impact the base chain?
techfeudalist 's avatar
techfeudalist 2 years ago
Could you please break that down a bit more for me? I assume you’re referring to the “silent”, miner coordinated soft fork. You refer to an attack. Is there a way that this can hurt the base chain? For example, if I ignore that this exists, could it hurt the incentives or my BTC in some way?
It's impossible for a softfork to be silent or only implemented by miners. Attempting to do so is just an attack, not a softfork at all. If Bitcoin users sit by and allow such an attack, it would be the end of Bitcoin.
RealJohnDoe's avatar
RealJohnDoe 2 years ago
EXACTLY! And this is what everyone must understand..⚖️👮‍♂️😠
RealJohnDoe's avatar
RealJohnDoe 2 years ago
What they don't realize too is that all that #inscriptions crap that's getting uploaded into the blockchain can be programmed to help attack the blockchain from within..😐
techfeudalist 's avatar
techfeudalist 2 years ago
I still don’t fully grok the attack. I can imagine you’re sick of the discussion (I saw you credited on the BIP website). Just hoping you’ll help me with more details. I’ve read the BIP writeup. To me, it sounds like a side protocol like ordinals / inscriptions. Old node software versions would see it as meaningless data that only means something to the participating users. Like I’m watching a baseball game and I see hand signs but don’t realize what they mean. Is it really an attack though? Does it improperly change network incentives? Censor transactions? Some other harm to the base chain? I get Lyn Alden’s point that we want the base chain to be resistant to change, but, unless I’m missing something, nothing appears to be changing from the perspective of old node software. It’s like a private group of people coordinating to use the blockchain to pass messages to each other. Again, appreciate any insight you can provide. 🙏
Thought you had said this in relation to BIP 300/301 on drivechains. My bad if the comment was totally unrelated. That’s what I was looking for.
RealJohnDoe's avatar
RealJohnDoe 2 years ago
Exactly! And that's why the artificially inflated #fees from non-Bitcoin financial transactions have to go. Along with #inscriptions like #ordinals #nfts #BRC20 etc. Including the proliferation of gigabytes of utxo fragments. It's all an attack on Bitcoin, and has already ripped off the owners and users of #Bitcoin for millions of dollars in criminal fees..⚖️👮‍♂️😠
techfeudalist 's avatar
techfeudalist 2 years ago
Interested to hear Luke’s thoughts on that thread. 👆 This drivechain system seems to a private group of miners passing messages to each other, storing data in the base chain. They seem to be following the “rules”. The drivechain proposal is overly complicated (with unknown long term consequences), but probably unstoppable if all current nodes will accept the data. Let’s hope this is benign because if following the protocol rules is an attack, then bitcoin is fundamentally broken.
It's an attack, not a softfork. And it can be recognized. *Worst case* it might require a PoW change to stop, but there's plenty of tools before that too
BIP300 and BIP301 go together. BIP300 has new txn type and the messages processed by miners and any nodes wanting to track state (anyone participating in a sidechain enabled by drivechain would be interested in this for fund state). Message types 1-4 are OP_RETURN. Types 5 and 6 are the deposit and withdraw tx. The tx are set as anyonecanspend, which necessitates economic majority nodes to enforce, not miners. If miners implement support and changes to consensus rules without adequate social consensus, this would constitute an attack. Layer 0 sets the rules, and consensus on drivechains hasnt been met yet
Scenario 1. No The economic nodes would be able to spend the funds intended to deposit or withdraw from the drivechains. Miners rejecting the spend txn would eventually lead to miners that would defect and spend themself, or other miners building on top of valid blocks. So maintaining 51% attack by miners means hoping for no defection and forgoing taking the funds themself and transaction fees (the funds are worth more) Scenario 2. Also no. Same as above, but miners are in a worse position. Non upgraded nodes would see the TX as anyonecanspend, and while thered be automation to spend, its really a miner that would either not be upgraded, or one that had support but defected who would win. All softforks carry a chainsplit risk, and the best way to avoid that is to either A) dont fork (status quo, where we are now), or B) ensure social consensus, then proceed with activation with BIP 8, LOT=true where miners can signal their readiness. Even after activation, I avoid putting funds at risk in new tx types for a time until im confident enough nodes truly are backing the new softfork rules to protect it.
So Paul has failed to realize that the reason hard forks are hard is because of consensus with other nodes. Now he thinks his BIP 300 group is enough to hard fork off the rest of the network because he doesn't think the rest of the network willl URSF, I feel he's dead fucking wrong and I'm putting my money where my mouth is. Miners are paid security. Don't follow the rules, don't get paid.
That's a 51% attack, since the transactions are legitimately anyone-can-spend, as enforced by virtually all full nodes. If the attack were interpreted as a softfork, that would not only mean miners get to dictate the protocol, it would also mean almost nobody runs a real full node (ie, enforcing ALL the rules). In other words, a fully centralized system.
@LukeDashjr what do you think about the practice of limiting the maximum length of the reorganization and disconnecting the node from the chain, is it useful or there are nuances?
Terry's avatar
Terry 2 years ago
Here’s a pressing question. Is that really you back on X?
As a jr pleb please help. Why won't state actors compromise BTC core devs and subtly alter things over time. Maybe in ways that are 4d chess and don't show as an attack vector till the 57th move? This protocol, plus this miner group, plus these wallets equal ??
Anchorite's avatar
Anchorite 2 years ago
This is why ossification is important. If it ain't broke, don't fix it.
dangershony's avatar
dangershony 2 years ago
What does it mean user enforced exactly? Spin 1000 nodes on a cloud do you have any power to enforce a softfork? A whale (or a few whales) on binance who never run a node can determine the value of the chain they prefer?