I selected three: bubble popping, bad news, and correction nothing to get all bent out of shape about.

Sounds like I'm contradicting myself, but that's really what I believe. The echo bubble popping was a necessary correction, but no reason to give up on Bitcoin or something. Just my opinion.

The final drops didn't look like panic selling, but like Bitcoinica executing liquidations of people who traded just this way. Surely there was panic selling as well, but the margin longs certainly amplified the crash. Face it, it was this exact trading strategy that brought its own downfall.

It's fairly evident in the price curve after it broke 4.8, that just didn't look like a typical panic-crash. Bitcoinica may have a better execution algorithm for the liquidations now, but it's still essentially the same effect.

Look, this is going nowhere. Yes, I'm not eager to write down the entire design right now, which is, as you expect, largely unfinished. And no, you can't bully me into doing that just yet. Actually, I'm fine with "is not good and thus I don't want to share it." It's true, in a way, since I just don't have any design documents that have the quality needed for sharing. An incomplete draft would just cause misunderstandings. Give it time. I'll gladly talk about things I believe to understand by now.

IMO, the real hurdle is that the discussion walks largely in unknown territory. The point is to figure out how to deal with anonymity and financial transactions at the same time, facing various types of attacks on both. All systems I know sacrifice one to allow safety in the other, even Bitcoin has no intrinsic anonymity outside of mining. To build a truly better TOR, it may be necessary to overcome this combination. We need a context of thinking that is aware of the dangers from both fields, it is rather about discipline in design than a single idea.

Let's talk about any concrete concept, and I'll contribute if I can. If this continues in a tone like "my psycho skillz show your thoughts are shit", chances are I just stop responding. In fact, I'm already angry at myself at formulating and editing a response to a post that mainly consists of insults.

Well, I have one for 90% of my Bitcoins. But, for one it changes all the time, and more importantly, I can't tell what other people would charge for their Bitcoins, or how many are lost exactly, etc. I can answer a lot about what I would do, but "all of Bitcoin" appears fairly illiquid to me, so who knows what would happen? The offer itself would value Bitcoin high enough for those who keep them, no? Hard to think in this kind of scenario.

It's not just that I'm not telling, I can't decipher the question in a way that gives a clear result. Either it's my current speculation price, which I won't tell of course and is probably below the lowest entry in the list, or it's some price that is affected by the fact that some large company decided something concerning Bitcoin -- which is too fuzzy in this formulation. that said, I wouldn't know how to answer if I wanted to.

That's not manipulation, that's actual damage dealt.

If people are serious, these fights are good. They make us strong. If people run off because they lost money in the glorious "buy BTC, get rich while watching TV" scheme... well, we can just hope there are other kinds of people around.

How much would I settle for so that someone can buy all Bitcoins? This makes no sense and should not be allowed according to the rules this economy agreed to work on. I refuse all offers that include anything but normal trading of exactly the coins I have.

Just to make this clear, I don't treat Bitcoin as a democracy. People do not have the right to sell "all of Bitcoin". If someone wants to buy every last coin, he has to do it normally, and this should be impossible.

If you want to have "just" the majority of my coins... try a rally right now, you might get them before double-digits. That's speculation for 'ya.

Sorry, but it wasn't my intention to layout our design here, but rather hint at the purpose of onion routing and the problems coming with it. Most "quick" proposals are too far off from our idea in terms of how the payments are processed to make me give a simple answer. I've discussed aspects, often on IRC with Bitcoiners, and found multiple threads on here that touch the topic, I'll try to not overlook things. I also talked to Bitcoiners from local meetups, looking for flaws and ideas.

To give a very rough overview, we plan to use a hybrid of Bitcoin and a custom system similar to RipplePay, which is based on yet another that does trust management. Allowing micropayments for traffic priced, but not immediately paid in BTC is one important aspect; distinguishing between trust of anonymity and trust for money another. Automating calculation of the latter type of trust may be the most important aspect, but might also be among the most difficult ones. We want to be truly decentralized.

Time is valuable, so I try to focus on the most relevant discussions. The proposals here are nowhere near matching our desired specification -- being dependent on the profitability of Bitcoin mining is way too risky to have me work on an implementation. Also, decentralization is not sufficient; the system could be disabled by targeting a few manually chosen mining pools. And nobody has explained the crypto that prevents information leaks with horrible dynamics, e.g. the relay knowing which pool which has issued the receipt. This results in a natural monopoly in which the largest pool offers the best anonymity, unless you build a really smart mixing system, which again poses problems not discussed here. And finally, the system is limited to users with hashpower, which is a massive limitation; paying in traffic would promote the pools to banks, adding a lot of manual trust management to the mix. No offense, but this is not the kind of quality we're aiming at, so solutions that would work in such a context are not really relevant for us. For us, this is all-or-nothing: we do it right or we fail trying.

In no way does that mean I'm opposed to any alternative ideas. Just working toward a slightly different goal.

Sorry to disappoint, but when people do such excessive volatility, especially caused by something as predictable as margin longs, I usually win as well. I don't even need to use short selling or the likes, I just assume that the price will be volatile enough.

There's no need to be a manipulator to profit from people changing their minds about price every few days. Let the price jump up again, and I will quote the post from last time a lot more often:

This doesn't mention Paxum. And add Tradehill to the list, it looks like they're seriously done for.

And the leverage mania is still maxed on longs. I'm really feeling better staying out of this.

Just barging in for a moment; been working on exactly this topic every now and then for the last few months.

I think it's possible, but the solution we're developing is very complicated. There are many problems in dealing with attacks on anonymity and attempts to scam the nodes. Very especially the problem is that, unlike in TOR, nodes will be optimized to maximize their own income, not the resulting network performance. This makes the protocol very tedious to design, since the resulting behavior must be proven to be the desired behavior.

Also, TOR itself is more complicated than it seems at first sight. Keeping the data flows anonymous and defending against the various means to censor such a network make every design a compromise. Taget China/Syria/Iran? Better have means to disguise the connection in various ways, and think of means to block probing. Running in northern Europe, not a bridge? It's suddenly all about performance and cheap traffic. And everywhere, there might be a user who wants to be very very certain to remain anonymous. One design that can adapt to all the circumstances would be great, but makes things even more complicated. Charging money adds to the trouble everywhere, think about questions like QoS and the resulting problems with congestion control, or how to behave if a node that has been paid does not deliver.

Extra fun: figuring out which node screwed up when more than one node in the circuit may be controlled by an attacker! Also annoying: if the target of a connection is on the "normal" internet, what happens if the server does not respond or responds with non-signed garbage? The exit node may have simply made up the response and demand payment for routing the "traffic".

Bottom line, I think it's an interesting problem, I want to work on it, however, "it won't be easy" is an understatement. Feel free to message me if you are interested in the project, we may need testers at some point. But don't expect anything to come up soon. Still playing with a variety of designs and probing for crypto problems, hard to tell where this is going.

Also, just a warning to anyone trying the same: don't do this the quick-and-dirty way. It would run great until someone finds an exploit in the dynamics, and then it might go down a lane it can't recover from. Prove that no client can systematically cheat the system to pay it, even if it has access to massive amounts of IP addresses, e.g. is a botnet or government. Similarly, prove that there is no cheap way to DoS the system to death using small amounts of money and again a large amount of IP addresses.

Magical excuse for running off with peoples' money: "We got SQL injected and the plaintext passwords were stolen!!"

It's so old by now, it's growing a beard. And Bitscalper reeked of scam from day one... seriously, what excuse did they have for staying anonymous again? What was their trading tactic, and why exactly wouldn't they trade with one investor's money instead of distributing the inflow?

I haven't heard a single credible explanation to Bitscalper's "business model" yet. I would assume that what's happening now is either a scammer screwing up because caring for security is too expensive, or an intentional leak with the purpose of covering the tracks at the cost of some of the BTC at hand.

e.g.: Limit withdrawals to 20%, leave gaping security hole, wait until hole is discovered, withdraw 80% to own BTC addresses; announce "we are sorry, we have been hacked, OMG a lot of money is gone!"

Just so I get this straight: I'm supposed to buy BTC now in the hope of selling them quickly afterward, to people who want to buy on Bitcoinica leverage at a higher price? Buy if the star is gone, sell if it's there?

Yea, excellent idea.

The behavior of traders lately really fed your tactic some extra fuel. Awesome to watch. Certainly beat me on the last week, I just didn't come up with a tactic and stayed conservative -- no losses, but hardly any profits. Mechanically draining the trend hypes probably is one very good tactic in such a situation.

I am reminded of that turtle from Love Hina that can fly at times. Good job on this.

Well, it shows 150%. What the hell does that mean?

Also, is "hedged" a part that is bucket shopped or a part that is traded on Mtgox?

;;bids number
;;asks number

Where number is a USD/BTC value in US formatting (e.g. 5.4). It will give the cumulative depth in BTC and USD.

Works in query and #Bitcoin channel.

btccharts.com shows completely different data though. Over 58k bids and over 34k asks. Clark Moody keeps drifting upward, it changes the accumulated orders when I refresh, and is somewhere in between otherwise...

Which sources are most reliable in such situations?

My impression is that btccharts and Clark Moody are generally not to be trusted when there's a lot of activity, and gribble usually gives credible data. But I can't know for certain, can I?

Edit: clark Moody seems to be in sync with gribble now if I refresh often enough. So I guess gribble was right again. Btccharts shows something different entirely.

Posting it here as well; sorry for double-post but traders might be interested. New user "Randall" just messaged me about the classic page working:

https://classic.mtgox.com/

It has the old "StartCom" signed key... doing a tracert shows that it is a different server: 72.52.5.81, also at prolexic, so I guess it's legit?

Randall just messaged me about a page that works:

https://classic.mtgox.com/

It has the old "StartCom" signed key... doing a tracert shows that it is a different server: 72.52.5.81, also at prolexic, so I guess it's legit?

WTF? A planned outage that shuts out only the EU, but keeps other people trading?

Genius! That's... I don't know... what a plan.

I see a route, but the site is down for me. Here is the route, starting from 6th hop:


  6    44 ms    45 ms    44 ms  if-7-2.tcore1.FNM-Frankfurt.as6453.net [195.219.50.2]
  7    41 ms    41 ms    41 ms  if-6-3.tcore1.AV2-Amsterdam.as6453.net [195.219.194.77]
  8    48 ms    41 ms    41 ms  if-2-2.tcore2.AV2-Amsterdam.as6453.net [195.219.194.6]
  9    49 ms    44 ms    44 ms  if-5-2.tcore2.L78-London.as6453.net [80.231.131.13]
 10    41 ms    41 ms    41 ms  if-2-2.tcore1.L78-London.as6453.net [80.231.131.2]
 11    41 ms    41 ms    41 ms  if-1-0-0-1600.mse1.LRT-London.as6453.net [80.231.130.42]
 12    38 ms    38 ms    38 ms  195.219.100.42
 13    38 ms    37 ms    38 ms  unknown.prolexic.com [209.200.156.70]
 14     *        *        *     Request timed out.
 15    40 ms    40 ms    40 ms  unknown.prolexic.com [209.200.156.73]
 16    38 ms    38 ms    38 ms  unknown.prolexic.com [209.200.156.34]
 17    39 ms    39 ms    39 ms  unknown.prolexic.com [72.52.5.67]

Trace complete.