The hack may be due to a hidden inflation bug. More LBTC were simply created, and a pegout was carried out legally.
Monero and Zcash are going to need a lot of luck.
Login to reply
Replies (22)
never say anything invincible in code
How do you know?
How do you know reality exists?
Yeah Quantum Computers can steal everyone's Bitcoin. Nothing is invincible. It's a fact that Monero hasn't had a hidden inflation bug before and there are factual reasons why that is the case: skill. Not luck.
The point I was making is that Monero has never had a hidden inflation bug that was patched on mainnet. That has never happened. The note I responded to suggests otherwise.
a little too cocky
But only if the supply increase is proportionally large. A few here and there wouldn't show up. That's what I would do if I want to sell them
The Liquid bug wasn't hidden inflation, the OP was incorrect. It was detectable. Monero has had detectable inflation bugs before, like Bitcoin.
Monero hasn't had a hidden inflation bug. For some reason people seem to think that's just random chance and aren't able to critically think why that fact is the case.
It's a hidden inflation bug.
That you don't understand it isn't my problem.
It was inflation of an amount you can't tell, but you can detect the transactions that exploited the bug.
If you want to call that "hidden inflation" I understand why.
It's completely distinct from a real hidden inflation bug where you actually cannot detect a valid transaction from a malicious one. Monero has never had such a bug patched in its history.
Just like in Bitcoin's overflow bug, you could detect the transactions exploiting the bug which allowed Bitcoin to roll back and erase that transaction from history.
That is also the case here, however this is a sidechain and the Bitcoin was already stolen.
Bitcoin's overflow bug would have led to this exact same consequence on Liquid.
how come maintainers not catch this before ? or keep open bounty for such things
Just so you know, the Liquid network could block the transactions that exploited this bug. The network can be rolled back to eliminate the inflation on the network because you can see which transactions exploited the bug, just like you could see which transaction exploited Bitcoin's overflow bug.
yes, correct. it was not technically hidden in the liquid case.
I assume they could've done it hidden, if they didn't peg out aggressively, or not at all. But there might be technical nuance that makes that impossible.
I also think the transaction on liquid didnt reveal it was fraudulent itself. it was only caught on liquid because there were unpatched nodes that saw the tx as invalid, so the network split.
Of course everything was very visible on the main chain
They couldn't have done it hidden. Every inflation transaction was detectable.
You can also say the Bitcoin overflow bug could have been done less aggressively by trying to cash out smaller amounts of Bitcoin at a time rather than a gazillion Bitcoin at once. The exploiter could have crafted transactions that were less conspicuous.
Are you sure? Could they have crafted a tx that doesn't peg out, just inflate and then slowly drain, or even just use it to transact inside of liquid? i don't have enough technical knowhow about liquid to know if this is possible, but if it's not possible, can you explain why?
I mean on bitcoin inflation would be very easy to spot, even if it's just 1 satoshi, since amounts are not obfuscated. there is no hidden way for inflation to happen on bitcoin
Every crafted transaction that causes inflation on Liquid has a proof attached to it that doesn't actually validate. You can detect inflation by validating all the proofs and seeing which ones fail.
The unpatched nodes split because they correctly validated proofs and identified the invalid ones.
The patched nodes were not using correct code to validate the proofs. They just need the correct code and would be able to reject all the invalid ones.
Hidden inflation is not that^. With hidden inflation, you cannot tell the difference between invalid transactions and valid transactions and so you can't go back and detect and erase invalid transactions like you can here.
It's true that confidential transactions increase complexity, but you can also break a sum tally in a way that "tricks" software via an overflow bug.
Right! Thanks for the explanation. ๐
because in this case, it relied on the cache, revalidation would expose the bad egg.
Agree, validation could be tricked through a bug, but it could still easily be traced if inputs and outputs don't add up.
A truly hidden inflation bug is only possible if amounts are hidden. would you agree?
(again, not saying it happend or it would happen, just talking about the possibility)
But what I'm saying is inputs and outputs can appear to add up to validation software if there is a bug like an overflow bug.
There are actually other ways for hidden inflation to happen even with plaintext amounts. I would agree that privacy features increase the surface area for potential hidden inflation bugs and that Monero has a wider surface for potential hidden inflation bugs than Bitcoin.
I think we mostly agree, and thanks again for the insights ๐ค
I guess the big main difference is with detectability of hidden inflation. If amounts are transparent, it would eventually be detected. if they are hidden, and the problem is in the proof system, it could stay hidden forever
is that a fair assumption?
I still don't think that's accurate. Bugs in software can take literally any shape. I think it's accurate to say hidden amounts make it more difficult to detect inflation bugs if they exist. That's a fair statement.
Fair! I'm happy to disagree with you on that one point.
I don't think the point you're making is wrong either, it souns solid to me.
Just have a different opinion on what is more/less probable
Here I will say that I can't think of a way for hidden inflation to happen on Bitcoin even theoretically. If you eliminate the linkable transaction graph, but kept plaintext amounts, then I can see how it's possible. So plaintext amounts + linkable transaction graph (everyone can see which inputs are destroyed in every transaction and which outputs are produced) = no theoretical hidden inflation. And that's Bitcoin.
That being said.. I think that can lead to a false sense of security. There are plenty of theoretical devastating inflation bugs or implementation bugs that would destroy Bitcoin (or Monero). Like broken cryptography that could allow anyone to steal people's Bitcoin.
I think it's definitely fair and valid to say Monero has a wider surface for these sorts of bugs (including hidden inflation which I can't see how Bitcoin could have), and therefore assign it a higher probability of occurring than Bitcoin. I agree with that.