I think that his points against transaction malleability are invalid:<p>- technical one - Bitcoin clients have a 100 ms delay before they relay messages. An attacker can compile a modified client that doesn't have these limitations and successfully outrun the rest. It was shown once that an attacker managed to successfully modify most of Bitcoin transactions on the network for some time in February<p>- social one - IIRC Gox had an automatic system, which reissued Bitcoin transfers if they failed. So you didn't need to phone them or convince in any way - Mt.Gox would send you a new transfer (and exhausting inputs has nothing to do here since they had no reason to use raw transactions API which lets you to use specific inputs, and instead they probably just used the more common sendto API) after it detected the old one failed (TXID not found on the network).
> <i>But elliptic curve crypto is not one of these topics. If the code can generate a handful of Bitcoin account numbers and corresponding keys correctly, there is hardly any reason why it cannot do so for all account numbers and corresponding keys.</i><p>Not totally true, not every input can yield a valid private key. The very upper ranges of the private key space are limited, as only integers 0x0 through 0xFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFEBAAEDCE6AF48A03BBFD25E8CD0364140 are valid private keys for Bitcoin.<p>You'd have to be stupid unlucky to randomly generate an invalid private key, but it can possibly happen.<p>> <i>If one must pick a cryptocurrency, the lowly dogecoin, of all things, is doing everything right. </i><p>Yeah, an ancient fork of Litecoin with a meme name is going to save us. Has absolutely no relevancy to the issue at hand of course.
<<The community has designated a Nobel leaurate as its nemesis, solely because he asked some inevitable questions every thinking person in his profession ought to ask>><p>If I'm not mistaken the Nobel leaurate [sic] in question wrote an article entitled "Bitcoin is evil." That seems to be slightly more than asking questions.
By the way have you seen Mark Karpeles public apology in Tokyo?<p>(20 seconds in) <a href="http://www.youtube.com/watch?v=15IZtzWOzRU" rel="nofollow">http://www.youtube.com/watch?v=15IZtzWOzRU</a><p>So he is French, educated in Paris and living in Japan since 2009?<p>Speaks French, English and Japanese. Sounds interesting, he's no dummy.
I think that there are only two real possibilities here: either Gox lost the money but doesn't know how they lost it, or they stole it. Theft is a <i>much simpler</i> hypothesis than many that are being proposed, but this doesn't really fit the pattern of the previous major thefts by wallets trusted by the community. The main difference is we that we know who these people are. It doesn't seem likely they could ever really cash-out without being observed. Even if they don't try to do that there are likely to be indictments and prosecutions that they will have to live through.
> The community has designated a Nobel leaurate as its nemesis, solely because he asked some inevitable questions every thinking person in his profession ought to ask.<p>Does someone know who he's referring to?<p>Edit: Thanks!
How about all the passport + proof of address data, required for registering with Mt.Gox. Where is it stored and has it been stolen / taken by third party? No one seems to ask any questions about this.
Question:<p>If the bitcoins were stolen, and the thieves later try to trade them, will that be obvious from the blockchain? Or can they successfully spend them without anyone realizing they are stolen?
We are talking about a half a billion dollar heist here. That's a lot of money - probably in the top 10 of biggest robberies ever committed.<p>You wouldn't have be a super hacker to pull it off. Some hidden cameras, USB key loggers and some microphones in the office could probably have gotten you a lot closer to that money.<p>And if you then could lure MtGox into emptying their hot wallet with the tx mal problem, then even better, but that was probably not even necessary.
If the CEO of MtGox Mark Karpeles is under gag order and he is on IRC, couldn't people confirm this by asking him while he is actively discussing some other topic on the channel, to publicly deny that he is under some sort of gag order. If he continues discussing other topics, without denying the gag order, it is an easy way for him to passively communicate that he is under such order without actually breaking the order.
I would think that insider theft is one of the least damaging outcomes for the Gox depositors.<p>Unfortunately I don't know that the Japanese government is going to have the technical expertise to properly identify the theft and track where the coins have moved. I can't imagine that the thieves have managed to squander all of the 750k BTC.
Of course this is wildly speculative but perhaps a simple answer is that someone internally at Mt Gox cleaned out the accounts and is blaming hackers and/or bugs. 100's of millions of dollars is easily enough of a temptation for someone to commit major fraud.
Btw, wouldn't it be easy to track down the mauled transactions and look who initiated them? After all, no one can use MtGox anonymously. Obviously, 'the hacker' could have used hacked accounts (this would have been noticed) or false identities.
I guess I don't see why the simplest explanation isn't that the US Feds seized the contents of the safe deposit boxes where their cold wallet was kept last year along with the $5m in bank deposits.
There are many interesting points made and dealt with in this article but what's weird/wrong/suspicious about a CEO using IRC?<p>Did he say something specifically stupid there? Or is the very medium tainted?
Far too sarcastic for something that is almost entirely raw, unsupported speculation. Further, it is conflicted -- it disbelieves some statements by Gox, while fully believing others (e.g. "they were in cold storage").<p>The one element that seems believable are questions about the malleability attack. I do not understand how Gox or any exchange or service wouldn't have an up to the minute, blockchain verified knowledge of exactly what their positions are. <i>Maybe</i> they only did such accounting weekly, or even monthly...but at some point over the supposed multi-year exploit they would have seen that account balances > address holdings.
Author put a lot of thought and work in to telling a great story, but...<p>Would be better if it weren't built on speculation, and limited by the things the author clearly doesn't understand about crypto.<p>Articles like this hurt the Crypto Currency movement because the things they get wrong about what did or didn't happen are speculation that just fuels fires of mistrust for what could happen. And the thing touted as solutions to it happening in the future aren't well researched so they give false security and opportunity for things to happen again.<p>I appreciate the authors effort to drive up the price of Dogecoin, and prevent further fall of BTC prices, but that's all this is.
The article is ok, and unsurprisingly did well on hn, but it's still the easy first level technical analysis. I learned nothing here.<p>I'd love to see a deeper analysis, but it probably can't come from a computer scientist.