Title: "Hash Chains" an idea for a new kind of "proof" Post by: CIYAM on July 09, 2014, 05:15:55 PM I have been playing around with a new idea for "proof" that might be suitable for a new kind of blockchain.
Blocks would be created by accounts whose hash matches closest to "hash chain" hashes published in previous blocks (yes - controlling the number of accounts that can be created is a key issue but one I have already considered). For those not familiar with what a "hash chain" is let me quickly illustrate (using MD5): MD5( "test" ) = 098f6bcd4621d373cade4e832627b4f6 MD5( "098f6bcd4621d373cade4e832627b4f6" ) = fb469d7ef430b0baf0cab6c436e70375 MD5( "fb469d7ef430b0baf0cab6c436e70375" ) = 25ab3b38f7afc116f18fa9821e44d561 so to create a new account a node published the hash: 25ab3b38f7afc116f18fa9821e44d561 When they "create their first block" they need to include the hash fb469d7ef430b0baf0cab6c436e70375 which can be easily checked to be correct from their initially published hash to be the "previous hash" (and cannot be reversed due to the fact that crypto hashes are irreversible). For their next block they would need to publish 098f6bcd4621d373cade4e832627b4f6 and they would "run out of the ability to create new blocks" once they have got this far. Apart from weak seeds (and MD5 for example purposes) can anyone explain any serious technical issues with this idea? Title: Re: "Hash Chains" an idea for a new kind of "proof" Post by: CIYAM on July 09, 2014, 06:00:11 PM This concept creates a "random lottery" to decide who creates the next block that is not biased due to *stake* and does not require the energy inefficiency that PoW has.
The only real issue is "account" creation and I think that can be solved by the creation of a "community driven" project (where accounts would have to be "earned" by performing "real work"). Title: Re: "Hash Chains" an idea for a new kind of "proof" Post by: franky1 on July 09, 2014, 06:23:27 PM This concept creates a "random lottery" to decide who creates the next block that is not biased due to *stake* and does not require the energy inefficiency that PoW has. The only real issue is "account" creation and I think that can be solved by the creation of a "community driven" project (where accounts would have to be "earned" by performing "real work"). now all i need to do is create 4096 addresses so that the first 2 characters cover everything from aa to 99, to have addresses that are as close to the 'lottery' hash of 2 difficulty at launch day.. then give me time to make 262144 addresses covering aaa -> 999 to get as close to the next lottery hash before your lottery tries to get harder to outsmart me with a difficulty of 3 characters matching.. (i might get lucky that one of the 262144 also has the 4th character that matches too) fastest cpu first, the game is on. who can get a close match to the 3rd blocks lottery hash you only need 16777216 addresses in your wallet to get as close to something that has the first 4 characters matching Title: Re: "Hash Chains" an idea for a new kind of "proof" Post by: CIYAM on July 09, 2014, 06:25:37 PM Yes - as stated the "account creation" would need to be controlled.
Apart from starting this blockchain as a "community project" I have several other ideas that would make it very difficult to acquire many accounts. So - it is only the "account creation" problem that you are seeing then? Title: Re: "Hash Chains" an idea for a new kind of "proof" Post by: franky1 on July 09, 2014, 06:33:03 PM Yes - as stated the "account creation" would need to be controlled. Apart from starting this blockchain as a "community project" I have several other ideas that would make it very difficult to acquire many accounts. So - it is only the "account creation" problem that you are seeing then? im in a joking mood tonight, it took me longer reading your initial post as i know your a long term bitcoiner, i was spending more time wondering if you were being serious at first.. so i started with a joke, but um.. yea controlling account creation.. as long as its code base not centralised based. which is a tougher one to get around because anyone can change the source to crack the code locally.. Title: Re: "Hash Chains" an idea for a new kind of "proof" Post by: CIYAM on July 09, 2014, 06:37:26 PM so i started with a joke, but um.. yea controlling account creation.. as long as its code base not centralised based. which is a tougher one to get around because anyone can change the source to crack the code locally.. Jokes are fine (but yes this is serious) - some basic rules in order to "create new accounts" would be that it would be "expensive to do so" (thus making it more "profitable at least in the short term" to just keep using the same account rather than *split it into two*). The very initial accounts would actually come from "working on the project itself" which could be via looking at github commits or other similar things (so the decentralisation grows from a project that is not initially decentralised). Where I am aiming this idea is a very different audience from the current "crypto-currency speculators" - I am thinking of of "blockchain apps" for things like franchise businesses (I also think it could be a great basis for a new kind of "social networking"). |