Quote from: fiatpete on November 17, 2014, 11:27:37 AM

Quote from: SapphireSpire on November 17, 2014, 03:59:20 AM

Quote from: Jungian on November 16, 2014, 09:16:12 PM

Quote from: SapphireSpire on November 16, 2014, 07:45:40 PM

Why do we need 3rd party users (miners) to confirm transactions and update the blockchain for us when anyone who has a valid copy of the blockchain can do it for themselves? By getting rid of 3rd party confirmations, transactions would be nearly instant and there would be no fees.

New coins can instead be generated as a % bonus to payees by an algorithm in the transaction process. Users could then 'mine' new coins by simply moving them between accounts and this appears exploitative at first but is necessary to compensate for deflation and create demand. Perhaps there are mechanisms that can minimize the consequences of abuse and keep it fair.

How would you pick a fork? Why not mine every fork then?

The same way we do now- by consensus. You have to accept the blocks that are reported by the largest number of nodes. The greater the number of honest nodes there are on the network, the more difficult it is to pull off a 51% attack but, without the 3rd-party confirmation scheme (mining), the risk is isolated to you alone rather than the whole network and only while you are syncing. And it doesn't waste energy.

It sounds like your arguing for a type of proof of stake based on number of nodes rather than longest block. Have a look at https://download.wpsoftware.net/bitcoin/alts.pdf for some criticisms of PoS from a bitcoin dev. You might also want to look at some of the PoS coins such as peercoin, blackcoin and nxt. If xcn hard forked to PoS now there would be an issue of how to fairly distribute the coins as the only way for new users to get any would be to buy off of existing holders.
Also for my node to know that the fork is reported by the largest number of nodes wouldn't it need to connect to alot more nodes than the usual 8 or 16 limit?

No, proof of stake is a 3rd-party confirmation scheme too. What I'm arguing is that there is no need for 3rd-party confirmation on new transactions because the blockchain enables all nodes to confirm new transactions for themselves and extend the blockchain. As long as all nodes begin with the same genesis block and follow the same rules for confirming new transactions, accidental forking shouldn't happen. Even the number and size of blocks between nodes can differ but they remain effectively synced as long as they share the same transaction history.

A consensus is only necessary for nodes with outdated transaction histories and this can only be done democratically. Old transaction data doesn't have to be downloaded all at once either. You don't need the data until someone tries to send you coins with references to transactions that you don't have a record of. You can echo those references over the network and, if valid, they should get confirmed by close to 100% of the network. If invalid, as part of a 51% attack, they should get rejected by close to 49% of the network. If you only trust transactions data with a very high % of confirmations, like 99%, you'll be pretty safe. Those transaction records are backed by older transaction records so, with every confirmed transaction, you can retrace the history of transactions in reverse.