FellowTraveler comments on Open-Transactions Auditing

opentransactions-auditing-comments

<FellowTraveler> okay amiller, et al, I believe I have finished writing up my thoughts / arguments on the auditing.
<FellowTraveler> There's an outline at the top, a list of scenarios, and a summary at the bottom.
<FellowTraveler> https://github.com/FellowTraveler/Open-Transactions/wiki/Auditing
<FellowTraveler> Basically I believe that amiller's requirement that third parties also have incentive to audit the transaction server, happens to be met in the case of Bitcoin, when OT is used in the scenario of voting groups.
<FellowTraveler> I believe that "kills two birds with one stone" and the rest of the article is a comparison of the various scenarios and their different incentives and vulnerabilities.
<FellowTraveler> I believe that without voting groups, there will not be any conventional Bitcoin issuers, since there is no value for them to offer in the realm of pure storage.
<FellowTraveler> Therefore I believe there will only be Bitcoin issuers who operate their own transaction server (such as knotwork) whom you must trust not to steal your coins, but whom cannot lie on your receipt.
<FellowTraveler> And there also will be Bitcoins issued on OT based on Voting Pools, where there will naturally be many other third-party auditors with incentive.
<FellowTraveler> (There will not be anything in between, where you have some Bitcoin issuer who is NOT part of a pool, and who does NOT operate his own transaction server — you can get this with gold, but not BTC IMO.)
<FellowTraveler> .
<FellowTraveler> I also believe that no crime is possible except for inflation, and that even that is not possible without keeping the receipts secret forever, since the crime is permanently detectable.
<FellowTraveler> I also believe that a gold issuer has no incentive to collude with a transaction server to commit inflation, and would bear the sole cost of doing so, to the sole benefit of the transaction server.
<knotwork> <FellowTraveler> Therefore I believe there will only be Bitcoin issuers who operate their own transaction server (such as knotwork) whom you must trust not to steal your coins, but whom cannot lie on your receipt.
<knotwork> makes sense to me, why would i let other people issue their crap on my server? except the otdemo server of course
<knotwork> if i am going to have tokens on my server claiming to represent grams of gold or litecoins or anything I want to be damn sure I know those things exist and where they are and that no one else can take them away
<knotwork> so to me its not really clear why one would even have a third party issuer for things like gold
<FellowTraveler> knotwork I believe there is a market for gold issuers because gold storage is a legitimate value-add that must be provided
<FellowTraveler> I believe that BTC storage is not a legitimate value-add.
<knotwork> instead of have the london gold vaults inc or whatever issue tokens on my server - which they are not likely to do it seems more likely i would have a gold vault account with them and i would issue tokens representing my gold i have in their vault
<FellowTraveler> BTC transactions on an OT transaction server, IS a legitimate value-add. But BTC storage is not.
<knotwork> if someone other than a huge lloyds of london insured gold vault wanted to isse gold on my server it makes sense to me to ask them to transfer the gold ownership to me then I will issue tokens
<FellowTraveler> the beautiful thing about OT is the market will be able to decide :)
<knotwork> if joe who has some gold in lloyds of londo vault and wants to issue tokens i think I would tell me issue your own damn tokens on your own damn server
<FellowTraveler> in the meantime I'm going to code the "confirm smart contract" opentxs command.
<knotwork> or use otdemo server where i am not responsible at all for what the fuck crap people issue there
<FellowTraveler> hmm I just updated the page at that link, there was a header missing.
<FellowTraveler> SUMMARY: If issuer operates his own transaction server, with gold or Bitcoins, you have to trust him not to steal the reserves. But he cannot lie on your receipt.
<FellowTraveler> If issuer and transaction server are separate entities, this will not ever actually happen with Bitcoin, but in the case of gold, there is no incentive for issuer to collude with inflation, which is the only possible crime, and which is only possible if the receipts are kept secret.
<FellowTraveler> If the currency is Bitcoin-based, and issued through a voting pool, then you have numerous third-party auditors and thus amiller's requirement is met.
<FellowTraveler> In all cases, no one can lie on your receipt, and inflation is the only possible crime, and there are entities with the incentive to perform audits to prevent that.
<FellowTraveler> with bitcoin voting pools you just end up with even more incentivized entities.
<FellowTraveler> There is still the possibility that a majority of the pool members could falsely vote on the blockchain to steal coins — just as a 51% attack on the blockchain itself could steal coins.
<FellowTraveler> But I believe in both cases, the is a definite security improvement over just trusting some asshole with your coins.
<FellowTraveler> And that's my argument, summed up, for those who don't have time to read the entire page.
<FellowTraveler> .
<FellowTraveler> Also, my ultimate conclusion is that risk can never be entirely eliminated, only managed and distributed. And that ultimately the responsibility for that, though all these systems are designed with that in mind, ends up in the wallet itself.
<FellowTraveler> In the sense that the wallet itself has to have functionality for the distribution and management of risk.
<FellowTraveler> At least, the smart wallet that the user of the future will want.