Topic: NXT :: descendant of Bitcoin - Updated Information - page 602. (Read 2761626 times)

I am doing a nearly atomic exchange of the actual crypto with NXT asset. The only glitch could be during an internal transfer, the NXT asset issuance fails for some reason. I havent had a chance to implement full error handling, it is only 10 days old, but when all the error cases are handled, all the confirmed crypto deposits will match up to confirmed NXT asset transfers.

The actual crypto is put into a multisig account. That is why it took me so long to get the alpha release. I had to tediously construct rawtransactions based on looking at all of the unspent outputs, which I had to recreate since bitcoind (at least dogecoind) didnt seem to properly track multisig transactions.

The AE asset therefore represents not just a promise to pay, but there will be a publicly visible multisig acct with the funds there. There will be a gatewaychain.info analogous to blockchain.info where you can verify deposits and withdrawals and their intermediate state.

To add more weight to the promise to redeem the asset for the crypto, multiple (3) gateways are simultaneously monitoring the NXT blockchain for asset transfers back that are coming in for redemption. All gateways query their local bitcoind and generate a proposed rawtransaction and a signed rawtransaction. Every single byte of the raw transaction must be identical between all of the gateways, then the selected gateway signs the already signed rawtransaction from one of the other gateways and submits it to the network.

I will be creating NODEcoin 2.0, which in addition to rewarding people for forging, will be actively validating all gateway transactions and will generate alerts if it ever detects any funny business. As long as the gateway monitors are happy, all is well. I dont expect more than an occasional yellow alert, when there is a network glitch and a transaction has to be manually resent. With the malleability issue, we need to be careful about blindly trusting any request for manual payment.

As long as the NXT community operated gateways continue to function and two of the independent gateway operators dont collude, all the withdrawals are assured. In case the deposit balance in the multisig accounts get too big, we can always fragment them and have dozens of multisig accts, each with a small fraction of the overall balance. Minimize the size of the acct to remove temptation to steal it, and require cooperation of another party to pull of the withdrawal.

I would think that most people would assess the security of this setup as much more secure than trusting some centralized exchange from going rogue or being hacked (c-cex for 300 BTC, poloniex for 150 BTC both this past week) or going belly up (Mtgox, etc).

Properly implemented after thorough testing and review, there would be no practical difference between trading a NXT AE crypto asset or the real thing. We will get the automated gateway so reliable that we will be able to offer NDIC (NXT deposit insurance), max 10000 NXT per person 100000 NXT total. So if the gateway ever causes a loss, there will be the insurance to cover some of the losses.

James

P.S. Can some marketing types take all this and make it nice and presentable?

CfB, is Nxt 90% attack-proof?

Locking this thread would be an absolutely terrible decision.  Aren't we all about the freedom to do what we want, why force people to a certain forum.  We are getting tons of stuff done here and other threads, leave it.

Everybody gets to vote.  I have edited my original post to clarify.

I had hoped to have this better organized at this point, but I've had a bad week.  Sorry.

Hey, guys, if you would, please edit your quote of my vote post so only the corrected original one shows up in the thread here.  Thx.

Nxt is mentioned here - http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2393940 (It Will Cost You Nothing to 'Kill' a Proof-of-Stake Crypto-Currency)

Resume of this paper: All PoS coins will die except Nxt coz it's 90% attack-proof.

I'm considering locking this thread on page 2100.  What are everyones' thoughts?

I don't want to interrupt communication, but I think we would make more progress if we were better organized.

Poll - https://forums.nxtcrypto.org/viewtopic.php?f=2&t=839

That is true, forging power will consolidate because people will want more consistent payouts.  So, limiting forging amounts per account is the only way to prevent colluding accounts from constructing a dangerously long chain.  Ian suggested limiting forging power to 1% or 10,000,000 Nxt.  I ran simulations with his program to give us some data to determine the best limiting forging percentage.  I ran each simulation three times.  All simulations were run for a period of ten years.  When I limited the forging power, I set the collusion rate to ninety percent.

If you do not limit forging power and set the simulator so one account controls 91% and nine accounts control 1% of Nxt, the longest blockchain the 91% holder can create is an average of 465 (514, 445, 435).

If you limit forging power to 1% per account and set the simulator for 100 accounts, the longest blockchain generated by the colluding accounts is an average of 142 (151, 137, 137).

If you limit forging power to 2% per account and set the simulator for 50 accounts, the longest blockchain generated by the colluding accounts is an average of 152 (166, 133, 156).

If you limit forging power to 4% per account and set the simulator for 25 accounts, the longest blockchain generated by the colluding accounts is an average of 1137 (1102, 1262, 1045).

If you limit forging power to 5% per account and set the simulator for 20 accounts, the longest blockchain generated by the colluding accounts is an average of 1452 (1443, 1459, 1454).

If you limit forging power to 10% per account and set the simulator for 10 accounts, the longest blockchain generated by the colluding accounts is an average of 1503 (1493, 1523, 1491).

From the results of the tests, we can see that limiting forging power to anything four percent or higher results in the colluding accounts being able to construct a longer blockchain than if we do not limit forging power and one account controls 90%.  Therefore, we should limit forging power to 1% or 2% per account.  I agree with Ian that forging power should be limited to one percent per account.

If we don't limit forging power, we have to accept the fact that we will have to wait for 465 minutes (one block per min) for confirmation if over ninety percent of the network is colluding.

I'm considering locking this thread on page 2100.  What are everyones' thoughts?

I don't want to interrupt communication, but I think we would make more progress if we were better organized.

Poll - https://forums.nxtcrypto.org/viewtopic.php?f=2&t=839

I don't think that we are at all. Firstly already only 1/4 of the "potential forging power" is being used (this will very likely change when people can grant forging rights to pools and we do not want to make it easy for 1 pool to have a huge amount of control which they *will* if there is no max. limit imposed).

Those that are "solo" forging are going to "split up" their NXT to ensure they get their maximum forging benefit (if they are bothered enough to forge then I think they would be bothered enough to do it in the most effective manner).

If you close this, people will just migrate back to https://bitcointalksearch.org/topic/ann-nxt-descendant-of-bitcoin-303898 

I'm considering locking this thread on page 2100.  What are everyones' thoughts?

I don't want to interrupt communication, but I think we would make more progress if we were better organized.

Poll - https://forums.nxtcrypto.org/viewtopic.php?f=2&t=839

A bit hard (meaning impossible) to know at any given point in time but you can analyse the historical data of the blockchain to work out which non-zero accounts have never forged or have not forged within a range of blocks (you can also add up the balances of all the accounts that *did* forge in that range).

When people can assign their forging right then I am guessing the total # of people forging will be *reduced* as some will opt out in favour of "easier money".

I'm considering locking this thread on page 2100.  What are everyones' thoughts?

I don't want to interrupt communication, but I think we would make more progress if we were better organized.

Poll - https://forums.nxtcrypto.org/viewtopic.php?f=2&t=839

Please tell us the details for future refrences on how to deal with them, thnx. and grats.

Do you know how to calculate what percentage is forging? If I can make each nodeminer do that calculation and also a check to make sure the acct is forging, we can create whatever reward we want. It could be per acct, per acct * sqrt(stake), per acct with max of X, per IP, etc.

The goal is to get people forging. If we make it in their financial interest to forge with a consistent generation of NODEcoin, then I think a lot of people will forge rather than not forge. Especially if the NODEcoins end up being worth more than the expected forging income, then the whole "pool" issue is moot.

BTER finaly gave me back my 100K Nxt. It took more than one month...

@james
(and others)

Speaking of Iceland...

I’m considering the marketing effect if we tested a NxtCoin as national currency in Christiania, Copenhagen. It’s a self-proclaimed autonomous neighbourhood of about 850 residents. But Christiania is visited by 500.000 tourists every year.

The media attention that Christiania gets, will benefit us. And vice versa. And in quite a few ways we share common grounds. It's like the old meets the new.

Christiania is a social experiment in it's origin. Talking decision making, it's a live version of our Nxt community. Chrisitanists make consensus-based decisions in matters that concern their whole community. OMG, it's long haired, and I admire it, the strength of discussing until consensus is reached. In difficult matters, this could be the next morning, or half a year. But they always get there.

Customers in Christiania choose whether they want their change in Danish Kroner or LØN (the Christianian currency since 1997). Christiania has it's own coin! - but they might be interested in going digital. Christiania has strong feelings about the environment, and with Nxt “making Greenpeace happy” and all, it would be a perfect fit.

Christiania "consensus"-meetings are closed for outsiders, but I am pretty sure that I can get in + bring another from the Nxt community with me, using my contacts. Even if the whole of Christiania does not agree on making a unique LØN NxtCoin their local currency, the shops in Christiania could easily begin accepting Nxt.

Imagine the media attention, if a society like Christiania agreed on making Nxt - or a coin made on top of Nxt - their "national currency". If "INTERESTED", let me know.

/apenzl