政策是规定不能作为支付手段的
但是你店小，成交额也小的话，估计也难被发现。
从性质上看，比特币应当是一种特定的虚拟商品，不具有与货币等同的法律地位，不能且不应作为货币在市场上流通使用。 "政策"只是說不能作為"貨幣", 但既然是商品, 就可以以物易物, 就像用一斤大米換兩瓶牛奶, 雙方同意就可以 |
|
|
|
Stop the FUD!!
Yes stop reporting news altogether. The news is true, but "OKCoin/Huobi become eBay-like platforms" is purely FUD |
|
|
|
|
200,000 BTC is what? It is the reward of 8,000 blocks. At this moment we have 296,825 blocks, and the 288826th block was generated in 2014-03-04, which is 48 days ago.
So we have generated 200,000 BTC out of thin air in the past 48 days. Have we crashed to $50? No.
|
|
|
|
|
1048XBT net buying yesterday. The total holding is at ATH of 102293XBT
|
|
|
|
|
我認為, *真正*會因為415問題失去信心而賣幣的人, 大部份已經賣了. 現在還會因為這些"新"聞賣的, 大都只是為了用更低價買回來
|
|
|
|
It would be nice to mine @Diff 1 with rigs from today  Hah  All blocks would be mine! :] if you had a 1TH/s machine in 2009, you would mine all the first 2016 Block really quick, then the difficulty would jump where it should be in order to get 10min average blocks rate..... I am not sure what is the maximum difficulty jump at once, I have to check that up. although you would have the majority of the network hashrate, you could do what ever you want assuming you know how to.... which makes me think about how lucky and thankful we are that Bitcoin is still alive today. Not exactly. The difficulty would jump only 4x at maximum, so that would be 1->4->16->64->......... until it's 10min average |
|
|
|
His private venture has invested in Bitpay |
|
|
|
So you just need a digital wallet to help you do all the accounting behind the scene, and generate a report for you at the end of the tax year
"just" I love that optimism We barely have wallets that don't randomly destroy your money or run on iOS... Yes, just IRS has nothing to do with insecure wallet or Apple policy, right? Uh, sure... IRS has nothing to do with this. Bolting complex accounting features on to existing apps, on the other hand, does. I'd say it is very simple. As all bitcoin transactions are timestamped on the blockchain, a software can trace all your addresses in the past year, match with the bitcoin price record, and calculate your capital gain/loss for each time you buy a cup of coffee. For the user, it literally requires only one click to generate the report. |
|
|
|
It could not be "all" because it is still random process, right?
Well, I meant 2/3 is a 51% attack, the miner can just orphan all the other miner's blocks.
Just assume all miners are honest
Anyway, it seems exponential distribution is the only choice
The key is that miners must be able to work together to build up the block. For example, you could update the header so that H(nonce1) < diff H(nonce1 | nonce2) < diff H(nonce1 | nonce2 | nonce3) < diff and so on. Once a miner finds a nonce1, they can broadcast it and then all the other miners can work on nonce2. You would need a way to reward the miners who find nonce2 though. An even easier way would be to increase the block rate to 1 block a minute, but only allow every tenth block to have transactions in it. Each block would still be allowed to have a coinbase though but it would be 10X times smaller. This gets the same effective block rate as now, but still means that miners can be rewarded for finding the (near) empty blocks. You could do this with a soft fork but you would need at least 91% of the network to make it work. A poll with only 10% of the hashing power would be able to create a fork that would be followed by older clients. Each would have to embed in the coinbase a reference to a chain containing 9 mini-blocks. Each mini block would have the same difficulty as the main chain block, but would be a header and a simple coinbase. When a new block is found, miners would switch to trying to build 9 mini-blocks on top of it. Once the nine have been found, they could build the next block. 90% of the tx + minting fees would have to be paid out to the mini-block's coinbases in order to keep things fair. This reduces the incentive to add transactions though. If you add a 1BTC fee transaction, you only get 0.1BTC. This may result in some pools not bothering with full blocks. System like this may work if the block reward is constant. People discovering miniblocks will get 1/10 of the reward. People discovering a main block will get 1/10 of the reward + transaction fee |
|
|
|
It would be nice if someone could show the maths. Let say we have 2 miners controlling 1/3 and 2/3 of total hashing power. What is the long-term block proportion in a 2-step mining?
2/3 of hashing power means all the blocks, no matter what. I ran a quick simulation with 100k blocks. hash_power = [0.3000 0.2000 0.1500 0.1000 0.1000 0.1000 0.0500] block_share = [0.4144 0.2250 0.1345 0.0685 0.0679 0.0698 0.0199] It could not be "all" because it is still random process, right? Anyway, it seems exponential distribution is the only choice |
|
|
|
And mining is no longer progress-free. Now the fastest miner finds all the blocks.
You are right, if we indefinitely increase the k (number of nonce requested) Although I can't show the maths but it could be shown with the following mental experiment. Let say I own 10 pieces of ASICs. Currently, no matter they are mining the same block or different blocks, my expected return is exactly the same. In my scheme, for the first step, the 10 ASICs may still mine independently. When a valid hash is found by one of the ASIC, the rest will stop first step mining and move to the second step immediately to increase the chance of getting a full block. This implies that efficiency of a fast miner is higher. It would be nice if someone could show the maths. Let say we have 2 miners controlling 1/3 and 2/3 of total hashing power. What is the long-term block proportion in a 2-step mining? EDIT: As the exponential distribution is the only memoryless continuous probability distribution, we have no choice but sticking to it? |
|
|
|
The block interval follows exponential distribution of lambda=1/10, with mean = 1/lambda = 10 and variance = square of the mean = 100. It is not uncommon to see block intervals of >1 hour and that could be really annoying. Having two blocks too close to each other is also a waste of block space. As the variance is the square of the mean, we may decrease the variance by decreasing the mean but that would have other side effects. It is possible to decrease the variance to 50 while keeping the mean as 10: 1. A miner will first mine the following message |Version|hashPrevBlcok|hashCoinbase|Bits|Nonce-a| until it is below the target. The difficulty is determined to archive an average successful rate of 5 minutes. hashCoinbase is the 256-bit hash of the desired reward transaction. The miner won't broadcast the message 2. When a valid hash is found in the step 2, the miner will start to hash the following message: |Version|hashPrevBlcok|hashCoinbase|Bits|Nonce-a|hashMerkleRoot|Time|Nonce-b| until it is below the target. The average successful rate is also 5 minutes. This is the valid block to broadcast to the network 3. A block is valid if it satisfies all the following requirements: - Hash of |Version|hashPrevBlcok|hashCoinbase|Bits|Nonce-a| is below the 5-minutes target
- Hash of |Version|hashPrevBlcok|hashCoinbase|Bits|Nonce-a|hashMerkleRoot|Time|Nonce-b| is below the same 5-minutes target
- Hash of the reward transaction is same as hashCoinbase
- All other existing rules, e.g. all transactions are valid, timestamp requirements
The block time interval will follow Erlang distribution ( http://en.wikipedia.org/wiki/Erlang_distribution) with lambda=1/5 and k=2. The mean is k/lambda=10, and variance=k/(lambda^2)=50 The block header size will be expanded from 80bytes to 116bytes. We can further decrease the variance by decreasing the difficulty while requesting more nonce. For example, we can decrease the difficulty to 1/3, then: First step: |Version|hashPrevBlcok|hashCoinbase|Bits|Nonce-a| Second step: |Version|hashPrevBlcok|hashCoinbase|Bits|Nonce-a|Nonce-b| Final step: |Version|hashPrevBlcok|hashCoinbase|Bits|Nonce-a|Nonce-b|hashMerkleRoot|Time|Nonce-c| The mean block time interval is still 10, but the variance decreases further to 33.33 For each additional nonce requested, 4 bytes additional bytes are needed Why need a hashCoinbase? We need to make sure people can't steal the partially completed block of other miner It may make some types of one-confirmation double spend attack easier. I'm not very sure. Any comments? |
|
|
|
So you just need a digital wallet to help you do all the accounting behind the scene, and generate a report for you at the end of the tax year
"just" I love that optimism We barely have wallets that don't randomly destroy your money or run on iOS... Yes, just IRS has nothing to do with insecure wallet or Apple policy, right? |
|
|
|
The current IRS guidance creates onerous reporting requirements on anyone using Bitcoin in a transaction, including retailers and consumers. This will negatively affect the adoption of Bitcoin as a currency, and thus negatively affect the inflow of new fiat. The linked news would remove this problem.
So you just need a digital wallet to help you do all the accounting behind the scene, and generate a report for you at the end of the tax year |
|
|
|
I'm pretty surprised to learn he still has a sizable amount there. There was plenty of chance to arbitrage and get out in the last bubble, even with profit. It was not sizable enough to warrant a flight to Japan. However, it is sizable enough to ruin my weekend. Loaded has disappeared since his weekend got ruined by gox. Hope he's alright |
|
|
|
Since I am also interested in the model, care someone tell, if the model in his opinion has predictive value, and does it give buy/sell signals or zones or anything?
y-axis is ln(price/trend)  It seems a bubble is likely to pop when the price is e^1.5 = 4.48 times over the trend line. Using 1.5 as cut-off....... in the first bubble, one would have sold at $17.16 on 04/06/2011, 5 days before the bubble pop but only 53% of the bubble top in the second bubble, one would have sold at $211.97 on 09/04/2013, just the day before the bubble pop with 79.7% of the bubble top in the latest bubble, one would have sold at $976 on 28/11/2013, 7 days before the bubble pop with 84.6% of the bubble top So the advice could be to start selling when it is over by 4.48 times. Sell more when it goes higher. There is no reliable buy signal from this plot |
|
|
|
如果你生成超过上限的私钥，则软件会覆盖掉较旧的私钥来腾出空间，这很可能会引发一个悲剧，因为覆盖后你将无法找回原先的私钥。
絕對沒有這回事, 應該是原文出錯了. coindesk上的文章亦已更正 |
|
|
|
这种是全手工操作，和当年的菜市场没什么区别，如果只能靠这种方法交易，成交量可能会是现在的千万分之一
這是因為你沒有想像力而已 這種系統除了可以買賣bitcoin, 也可以擴展到買賣交易平台的現金碼 (btcc, okcoin, btce都有提供了) 如果你需要快速交易, 可以經這類p2p平台買入bitcoin或現金碼, 就可以充值到傳統平台 如果你需要把bitcoin換為真正的人民幣, 或把傳統平台上的法幣換為真正的人民幣, 都可以經過p2p平台賣出 有買有賣, 市場就形成了 |
|
|
|
|
Show Posts
