Are instantly confirmed decentralized transactions even theoretically possible?

[monsterer]

Thank you for the math, but what about the idea of providing the proof that coins weren't spent and won't be doublespent? Depending on your definition of "decentralized" I might describe a system that will work.

What's to stop the double spender sending the same proof to both parties involved in the attempted double spend?

[Come-from-Beyond]

What's to stop the double spender sending the same proof to both parties involved in the attempted double spend?

Cryptography.

[Peachy]

CfB, are you referring to the idea of using Quantum Entanglement (e.g. spooky action at a distance)?

That would be some off-the-chain next-level crypto.   

http://phys.org/news/2015-03-quantum-einstein-spooky-action-distance.html

[americanpegasus]

Barring some major breakthrough, it's also a decade or more away.   

[Fuserleer]

Why no stargates mentioned in this thread? 

[onemorexmr]

Why no stargates mentioned in this thread? 

it takes too long to open them. mine usually takes around 5secs to open...its faster to send through a network

[Fuserleer]

Why no stargates mentioned in this thread? 

it takes too long to open them. mine usually takes around 5secs to open...its faster to send through a network

hahaha touche!

[amaclin]

Are instantly confirmed decentralized transactions even theoretically possible?

No.
https://en.wikipedia.org/wiki/CAP_theorem

[herzmeister]

https://en.wikipedia.org/wiki/CAP_theorem

Practically all modern high-scale web sites and services let go of the "1) Consistency".

https://en.wikipedia.org/wiki/Eventual_consistency is the keyword these days. It's apparently good enough. Bitcoin uses it almost by definition. The question then becomes how it can be optimized and how fast it can be designed.

[amaclin]

The problem is that "confirmed" != "secured".
There is not a problem to have blocks each second (instead of 6 per hour in current bitcoin)
And the transactions will be confirmed "instant", in a couple of seconds.
But one, two and even 10 confirmations in this crypto will be unsecure.

[monsterer]

The problem is that "confirmed" != "secured".
There is not a problem to have blocks each second (instead of 6 per hour in current bitcoin)
And the transactions will be confirmed "instant", in a couple of seconds.
But one, two and even 10 confirmations in this crypto will be unsecure.

Actually, below 10 seconds block time you start to hit big inefficiency problems due to orphans - you get around 200 orphans/day at 10 seconds, even with stuff like GHOST. Bitcoin has about 1 orphan/day.

[franky1]

the way that instant confirms would work is not to speed up the block solving. but to instead lock-in a TX into the miners mempool. and ignore any second tx's using the same funds. even if the second one has a higher tx fee. and have all miners advertise their mempools so that everyone knows and all miners have the same locked in tx's. that way double spending (attempts) wont be a worry, even with zero confirms

[amaclin]

that everyone knows and all miners have the same locked in tx's. that way double spending (attempts) wont be a worry, even with zero confirms

Where do you live? Is that a world where the speed of the light is infinite, network latency is zero and the connections are permanent?  

[monsterer]

the way that instant confirms would work is not to speed up the block solving. but to instead lock-in a TX into the miners mempool. and ignore any second tx's using the same funds. even if the second one has a higher tx fee. and have all miners advertise their mempools so that everyone knows and all miners have the same locked in tx's. that way double spending (attempts) wont be a worry, even with zero confirms

Nope. That is one poor attempt at solving the problem - here is why that doesn't work:

1. You would need to hold all transactions ever made in the mempool

- otherwise later blocks can still reverse transactions

2. Sybil attack - transaction locks can be faked if you control a majority of nodes

- your sybil nodes confirm a lock for transaction A and then also confirm another lock for a double spend of A after the merchant accepts your initial transaction. If there is no cost to being a node, this attack is free. If you must supply collateral to be a 'node' then the cost is a constant.

[monsterer]

BitShares does 3 second transactions currently at speeds of up to 1000tps and still improving...  Not bad for now.  

https://bitsharestalk.org/index.php/topic,18727.msg240918.html#msg240918

You might think that this post belongs in the altcoin section, but bitcoin will hardfork some of this code in the future, so here it stays.

Transactions in 2.0 might get into a block in 3 seconds, but it takes 30s for them to become 'irreversible'.

[amaclin]

Transactions in 2.0 might get into a block in 3 seconds, but it takes 30s for them to become 'irreversible'.

Nothing is 100% irreversible.

[monsterer]

Transactions in 2.0 might get into a block in 3 seconds, but it takes 30s for them to become 'irreversible'.

Nothing is 100% irreversible.

Hence my quotes around the word.

[amaclin]

Hence my quotes around the word.

chopped meat can not be reversed to a cow 

[Come-from-Beyond]

chopped meat can not be reversed to a cow 

Not sure... http://www.livescience.com/49610-scientists-unboil-egg.html

[TeamButtcoin]

just because entangled particles do things simultaneously does not mean in any way that 'quantum computing' will allow for instantaneous transmission of or processing of data, neither are possible