Topic: NXT :: descendant of Bitcoin - Updated Information - page 548. (Read 2761624 times)

well boys, gear up!  if you thought the excitement and payouts at blocks 22001 - 22005 were big, just wait till AE gets turned on....

Done

voted.

Guys, here needs help:

http://cryptocrimson.com/poll-which-alt-coin-will-see-the-most-growth-in-2014.html

Guys, here needs help:

http://cryptocrimson.com/poll-which-alt-coin-will-see-the-most-growth-in-2014.html

In order to have the NXT core directly interface with the BTC blockchain, we would need to put some bitcoind functionality into the NXT core. That would mean finding a Java implementation or porting bitcoind, and each and every update to bitcoind. So, it is not impossible, just impractical. By the time the new bitcoind version is ported to Java, a new one is probably just around the corner.

In addition to being impractical, doing a cross chain transaction for each transaction adds a lot of overhead and time, especially if somebody is actively trading. There is also no assurance that the trade will complete since it requires both parties to have fully funded account and to follow through with sending their side of the transaction.

The advantage of NXT blockchain <-> NXT AE <-> BTC blockchain, is that the deposit of BTC is automated and when it is traded in AE, there is a guarantee of payment. So while technically it might not be automated cross chain, it actually solves the problem of automatically converting NXT <-> BTC.

I am not the expert in the secret sharing methods for the atomic cross chain trades, so it might be possible to add to the NXTcore the required secret sharing logic that is needed for this. Then there would need to be a way for the two parties to come to agreement on price and we cannot use the AE, so that is essentially the same sort of thing, just different. After the two parties agree to a price, they would need to invoke some sort of semi-automated process where each one locally signs and broadcasts their half of the shared secret transaction. There is no assurance that both parties will fully satisfy the transaction and it could all just not happen if either of them bails out. Then assuming we have ported a subset of bitcoind into the NXTcore, it could do the rawtransaction secret sharing monitoring.

It just seems to me that the NXT blockchain <-> NXT AE <-> BTC blockchain solution is much closer to what users are expecting. It also has the advantage that people will be able to trade any crypto against any other using a decentralize AE. That will be a uniquely powerful advantage for NXT.

Which solution to automated crypto trading is your preference?

James

Maybe this qualifies for the extra 2 million NXT for extra features?

Thanks!

I am doing my best to reduce the trust required for the gateways. I think if we had 100 gateway servers, all doing peer review, along with thousands of nodecoin clients, then even though ultimately the gateway transaction is 2 of 3 multisig, it really has a lot of built in decentralization and minimal trust is required.

Maybe this qualifies for the extra 2 million NXT for extra features?

Anyway, I am going full speed ahead with the multigateway solution. It also happens to solve the NXTcash deployment problem. Basically with 100 gateways, it creates a mini-network of its own and since it only has to do a few things, we can harden it against wide open attacks. The 100 servers would act as guardians for the three servers that do the actual gateway money handling. Only the 100 guardian servers would even be allowed to communicate with the three gateways.

James