I wanted to relay an interesting related link that Melvin PMed me:
https://petertodd.org/2016/commitments-and-single-use-seals
@Aronesty
Thanks, that system looks interesting, I'll have a closer look!
@Voskuil
I think we must disagree on at least one fundamental point. I'm finding myself disagreeing with most of what you're saying.
> If perfect is not possible, it’s not possible. It reduces to trust, which is the status quo.
Let not perfect be the enemy of good. Trust cannot be eliminated. Perfection cannot be achieved. However trust can be reduced further than it exists today, and we can clearly make things better than they are now. Are you really saying that if its not perfect, its worthless?
> All “users” need to simultaneously share their individual and temporary audits with each other (ie publicly).
This is not the case. In both the mechanism I briefly described and Peter Todd's mechanism from Melvin's link (top of this message), users need not share any information with other users unless that information is "my balance doesn't match the record". It doesn't even need to be in a timely manner if these records are committed to the blockchain - one could theoretically look back years and compare their personal records to the records published by the custodian, and then tell the community about it if they don't match. All that is required is that a critical mass of users verify their balance (ideally using software that regularly checks automatically).
> It is not hard to spot price inflation
I don't know why you think that's the case. Inflation is today vehemently argued about. Are high real estate prices indications of inflation? What about recovering stock prices? Is inflation temporary or will we expect it to last more long term? If one company is inflating perceived supply, this would almost certainly not show up as significant economy-wide inflation. And if the majority of companies are doing it, how can you stop it if you don't have any idea which ones are doing it? I do think spotting the inflation in any kind of timely manner is indeed hard.
> Stopping or avoiding it is the actual issue. No “proof” of reserve can do this.
I of course agree that proof of reserves cannot stop inflation/insolvency. I disagree with the idea that this is the "actual" issue - which seems to imply to me that its the only issue that matters. Even if we can't stop a company from promising more coins than they have in reserves, we can limit how long these events happen for - and how big these bubbles can get. Don't you agree that would be very helpful, if it were it possible (which it seems you think it isn't)?
> The federal reserve was clearly insolvent from its early days, as that was its purpose.
Do you have a source as evidence that it was widely understood that the Fed was insolvent from its early days? I really don't think it was seen as the purpose by the vast majority of people in the US. People were lead to believe every dollar was backed by a unique chunk of gold. Had it been possible to do a kind of proof of reserves (or estimate of reserves) as we're talking about, it would have been clear much earlier that the Fed was doing a lot of shinanigans. I think that would have been very useful information for the public back then. Perhaps they wouldn't have been able to do anything about the Fed (or maybe they would have). But people can certainly pull their money out of companies that can't show solvency.
> Nonsense, any business can fail, regardless of temporal cash reserves.
I agree that any business can fail. But a bank that pretends it can serve cash on demand is not a normal business, and cash reserves absolutely relate to their ability to survive as a bank. Its honestly confusing to me how you could think otherwise. Also, calling my thoughts "nonsense" is rude, please check yourself, Eric.
> it's hardly an improvement over holding your own keys.
No one is claiming that proof of reserves is "an improvement" over holding your own keys. Clearly holding your own keys is ideal. However, not everyone is comfortable with that. The fact of the matter is that many will choose a custodial wallet, for better or worse. PoR attempts to make thing on the "better" side than the "worse" side.