5 ms·

The damage is done. The contract is not immutable. Squirting UDP packets until you see your tx pop up in a block turned out to be a risky design choice. Hope

The damage is done. The contract is not immutable. Squirting UDP packets until you see your tx pop up in a block turned out to be a risky design choice. Hopefully, Solana can add p2p tx propagation like a normal public blockchain. I just wish they wouldn’t throttle by stake as that makes Solana more of a private hyperledger than an Ethereum replacement.

saurik 4y ago
But we knew it explicitly wasn't even intended to be immutable: Solend had "governance", which has the explicit goal of allowing for decentralized decisions to upgrade the contract and change the rules of engagement. This is very much unlike when ecosystems like Ethereum or Solana decide to modify contracts that were intended to be immutable because the entire ecosystem kind of shifts with the decisions of the developers without really questioning it; this was an explicit vote taken by the people who own the governance token for Solend, a vote you yourself could have participated in... which is the real problem, as far as I'm concerned: there was a 1% quorum requirement on that vote, which they barely made, mostly due to one larger participant! This is a democracy -- which, to be clear, is very much "decentralized" -- but designed in a way where not only does it have the classic issue of how they "can only exist until the majority discovers it can vote itself largess out of the public treasury" but additionally has the problem that our real-world governments often do (and at least the one I am a part of here in the United States is well known for): that so few people vote that in practice decisions are made by minorities.
- mcintyre1994 4y ago
  > Solend had "governance", which has the explicit goal of allowing for decentralized decisions to upgrade the contract and change the rules of engagement. I’m not sure if this was true before 24 hours ago? If you click back on the link of this post you get to the DAO page: https://realms.today/dao/SLND https://realms.today/dao/SLND It seems that the vote to take control of the whale account was the first ever vote of this DAO. The Solend twitter account also refers to it as the first governance vote: https://twitter.com/solendprotocol/status/1538441350441422848 https://twitter.com/solendprotocol/status/153844135044142284... I’m not sure exactly when the Solend DAO was formed, but it doesn’t seem true that past users were involving themselves with a Solend that had governance. That seems to be an innovation specifically to solve this problem.
  - gruez 4y ago
    >I’m not sure exactly when the Solend DAO was formed, but it doesn’t seem true that past users were involving themselves with a Solend that had governance. That seems to be an innovation specifically to solve this problem. A quick search for "site:solend.fi governance" kinda disproves this. There aren't any explict pages on how governance works, but there are pages[1][2] (last modified 2 months ago) that havementions to governance tokens. [1] https://docs.solend.fi/Architecture/access-controls https://docs.solend.fi/Architecture/access-controls [2] https://docs.solend.fi/protocol/bug-bounty https://docs.solend.fi/protocol/bug-bounty
    - mcintyre1994 4y ago
      Kind of. The first link has: > For now, there isn't a timelock program, but transitioning to governance in the future will cause upgrades to go through a governance vote as well. So not really suggesting any currently functional governance, which makes sense given they never did a vote before. That said the important thing here is probably that they document an upgrade authority, and that they control its keys. The second link mentions governance only in the context of types of attacks that are in scope for a bug bounty, which doesn’t really tell users anything about governance. I wouldn’t be surprised if they copied this from somewhere TBH, it doesn’t seem very tailored.
artdigital 4y ago
Isn't that exactly what's currently happening with the transition from UDP to QUIC?
- favflam 4y ago
  Yes, with stake-weighted tx bandwidth. I think if QUIC (node to node handshake) was in place, Solend would have no excuse to do this vote as there would be very little risk of the network going down.
rvz 4y ago
The DAO hack of 2016 and the decision to go back a few blocks and revert the hack by a few at the top already taught us about the nonsense of 'cOdE iS lAw' years ago and that decentralization being a total lie. The move to proof-of-stake will only show that it will get worse.
- quickthrower2 4y ago
  I think they blocked the attacking address in code, rather than going back blocks (which would revert everyone elses txns. Doesn’t affect your point tho.
- TacticalCoder 4y ago
  > The DAO hack of 2016 and the decision to go back a few blocks and revert the hack by a few at the top already taught us about the nonsense of 'cOdE iS lAw' years ago and that decentralization being a total lie. There's been zero rollback to block the attacker's seized ETH in the DAO hack. There was a "cooldown" period during which the attacker couldn't withdraw the fund and during that cooldown period Ethereum's code was modified and everybody had the choice to either run the old code or the new code on its node. People mostly decided to adopt the code preventing the attacker from accessing the stolen ETH while some decided to continue with the original code. So the chain forked into two blockchains: Ethereum and the one that was named "Ethereum Classic" (ETC). Arguably after the fork ETC was the "original" blockchain (but with its token way less valued) but since then AFAICT the real development has been happening on Ethereum, not Ethereum Classic and so nobody is arguing anymore that Ethereum Classic is the "real Ethereum". > The move to proof-of-stake will only show that it will get worse. Why would that be? Instead of rich miners it's now rich HODLers mining blocks but mining is ultra centralized through a few big pools: I don't see how PoS, if it works, would make anything worse than it currently is?
- ricochet11 4y ago
  > The move to proof-of-stake will only show that it will get worse. why? seems like an unrelated statement to your first line.
lima 4y ago
> The damage is done. The contract is not immutable. It never was immutable, and generally, that's fine - contract upgrades by DAO vote are pretty normal and hard to avoid for complex projects. On Ethereum, you need a proxy contract to implement it, on Solana, there's a native loader that can set an upgrade authority and which works similarly. Either way, working as advertised. What appears to be utterly broken is Solend's DAO governance process, which apparently allows a single whale account to approve changes. There is nothing decentralized about that! > Squirting UDP packets until you see your tx pop up in a block turned out to be a risky design choice. Yes, but how is that related to Solend's (!= Solana) governance? (Solana is fixing this by moving the transaction ingress stage to QUIC with proper flow control - it won't be throttling by stake, but have guaranteed minimum allocations for staked nodes)
- YPCrumble 4y ago
  I’m not very well versed in crypto so I apologize if this is off-base. Are you saying that “working as advertised” means that a crypto blockchain is adjustable by a centralized authority?
  - wilonth 4y ago
    Solana programs have a "final" flag that could be turned on to make it immutable. This project doesn't have it turned on so it's open to update.
    - weego 4y ago
      In short, yes.
      - lima 4y ago
        Actually... no. Solana has no central authorities. Solend is an individual application running on Solana.
  - im3w1l 4y ago
    The blockchain itself remained append only. No authority went in and rewrote history. A contract running on top of the blockchain was updated in accordance with its own predefined governance rules. This is an advertised feature of the blockchain. Those governance rules the contract chose were arguably too loose.

saurik 4y ago

But we knew it explicitly wasn't even intended to be immutable: Solend had "governance", which has the explicit goal of allowing for decentralized decisions to upgrade the contract and change the rules of engagement. This is very much unlike when ecosystems like Ethereum or Solana decide to modify contracts that were intended to be immutable because the entire ecosystem kind of shifts with the decisions of the developers without really questioning it; this was an explicit vote taken by the people who own the governance token for Solend, a vote you yourself could have participated in... which is the real problem, as far as I'm concerned: there was a 1% quorum requirement on that vote, which they barely made, mostly due to one larger participant! This is a democracy -- which, to be clear, is very much "decentralized" -- but designed in a way where not only does it have the classic issue of how they "can only exist until the majority discovers it can vote itself largess out of the public treasury" but additionally has the problem that our real-world governments often do (and at least the one I am a part of here in the United States is well known for): that so few people vote that in practice decisions are made by minorities.

mcintyre1994 4y ago
> Solend had "governance", which has the explicit goal of allowing for decentralized decisions to upgrade the contract and change the rules of engagement. I’m not sure if this was true before 24 hours ago? If you click back on the link of this post you get to the DAO page: https://realms.today/dao/SLND https://realms.today/dao/SLND It seems that the vote to take control of the whale account was the first ever vote of this DAO. The Solend twitter account also refers to it as the first governance vote: https://twitter.com/solendprotocol/status/1538441350441422848 https://twitter.com/solendprotocol/status/153844135044142284... I’m not sure exactly when the Solend DAO was formed, but it doesn’t seem true that past users were involving themselves with a Solend that had governance. That seems to be an innovation specifically to solve this problem.
- gruez 4y ago
  >I’m not sure exactly when the Solend DAO was formed, but it doesn’t seem true that past users were involving themselves with a Solend that had governance. That seems to be an innovation specifically to solve this problem. A quick search for "site:solend.fi governance" kinda disproves this. There aren't any explict pages on how governance works, but there are pages[1][2] (last modified 2 months ago) that havementions to governance tokens. [1] https://docs.solend.fi/Architecture/access-controls https://docs.solend.fi/Architecture/access-controls [2] https://docs.solend.fi/protocol/bug-bounty https://docs.solend.fi/protocol/bug-bounty
  - mcintyre1994 4y ago
    Kind of. The first link has: > For now, there isn't a timelock program, but transitioning to governance in the future will cause upgrades to go through a governance vote as well. So not really suggesting any currently functional governance, which makes sense given they never did a vote before. That said the important thing here is probably that they document an upgrade authority, and that they control its keys. The second link mentions governance only in the context of types of attacks that are in scope for a bug bounty, which doesn’t really tell users anything about governance. I wouldn’t be surprised if they copied this from somewhere TBH, it doesn’t seem very tailored.

artdigital 4y ago

Isn't that exactly what's currently happening with the transition from UDP to QUIC?

favflam 4y ago
Yes, with stake-weighted tx bandwidth. I think if QUIC (node to node handshake) was in place, Solend would have no excuse to do this vote as there would be very little risk of the network going down.

rvz 4y ago

The DAO hack of 2016 and the decision to go back a few blocks and revert the hack by a few at the top already taught us about the nonsense of 'cOdE iS lAw' years ago and that decentralization being a total lie. The move to proof-of-stake will only show that it will get worse.

quickthrower2 4y ago
I think they blocked the attacking address in code, rather than going back blocks (which would revert everyone elses txns. Doesn’t affect your point tho.
TacticalCoder 4y ago
> The DAO hack of 2016 and the decision to go back a few blocks and revert the hack by a few at the top already taught us about the nonsense of 'cOdE iS lAw' years ago and that decentralization being a total lie. There's been zero rollback to block the attacker's seized ETH in the DAO hack. There was a "cooldown" period during which the attacker couldn't withdraw the fund and during that cooldown period Ethereum's code was modified and everybody had the choice to either run the old code or the new code on its node. People mostly decided to adopt the code preventing the attacker from accessing the stolen ETH while some decided to continue with the original code. So the chain forked into two blockchains: Ethereum and the one that was named "Ethereum Classic" (ETC). Arguably after the fork ETC was the "original" blockchain (but with its token way less valued) but since then AFAICT the real development has been happening on Ethereum, not Ethereum Classic and so nobody is arguing anymore that Ethereum Classic is the "real Ethereum". > The move to proof-of-stake will only show that it will get worse. Why would that be? Instead of rich miners it's now rich HODLers mining blocks but mining is ultra centralized through a few big pools: I don't see how PoS, if it works, would make anything worse than it currently is?
ricochet11 4y ago
> The move to proof-of-stake will only show that it will get worse. why? seems like an unrelated statement to your first line.

lima 4y ago

> The damage is done. The contract is not immutable. It never was immutable, and generally, that's fine - contract upgrades by DAO vote are pretty normal and hard to avoid for complex projects. On Ethereum, you need a proxy contract to implement it, on Solana, there's a native loader that can set an upgrade authority and which works similarly. Either way, working as advertised. What appears to be utterly broken is Solend's DAO governance process, which apparently allows a single whale account to approve changes. There is nothing decentralized about that! > Squirting UDP packets until you see your tx pop up in a block turned out to be a risky design choice. Yes, but how is that related to Solend's (!= Solana) governance? (Solana is fixing this by moving the transaction ingress stage to QUIC with proper flow control - it won't be throttling by stake, but have guaranteed minimum allocations for staked nodes)

YPCrumble 4y ago
I’m not very well versed in crypto so I apologize if this is off-base. Are you saying that “working as advertised” means that a crypto blockchain is adjustable by a centralized authority?
- wilonth 4y ago
  Solana programs have a "final" flag that could be turned on to make it immutable. This project doesn't have it turned on so it's open to update.
  - weego 4y ago
    In short, yes.
    - lima 4y ago
      Actually... no. Solana has no central authorities. Solend is an individual application running on Solana.
- im3w1l 4y ago
  The blockchain itself remained append only. No authority went in and rewrote history. A contract running on top of the blockchain was updated in accordance with its own predefined governance rules. This is an advertised feature of the blockchain. Those governance rules the contract chose were arguably too loose.