proof of work - better calculation - Githubissues

zold-io / papers

Zerocracy and Zold official PDF documents

https://papers.zold.io

9 stars 3 forks source link

proof of work - better calculation #11

Open ghost opened 6 years ago

ghost commented 6 years ago

Use a better proof of work process to make a more meaninful work.

Also the score should use more parameters to be more specific

0crat commented 6 years ago

@yegor256/z please, pay attention to this issue

0crat commented 6 years ago

@oltdaniel/z this project will fix the problem faster if you donate a few dollars to it; just click here and pay via Stripe, it's very fast, convenient and appreciated; thanks a lot!

yegor256 commented 6 years ago

@oltdaniel I'm not sure I follow. What exactly are you suggesting?

ghost commented 6 years ago

@yegor256 We talked about to choose a better proof of work algorithm to stop this meaningless iteration of hashes.

yegor256 commented 6 years ago

@oltdaniel what would be a better option? I'm open for suggestions.

ghost commented 6 years ago

@yegor256 The score is currently based on the hashes a node calculated. These hashes are generated by generating a new SHA256 hash that has :strength: leading zeros. This results in an new alias for the score, computing power per node. To make this a more meaningful calculation, we could let the nodes hash the wallets over an over, and set a score based on that, which will lead to the same result (yes, this will require a new formula) and represents the computing power per node as well as the wallet count parameter in one score.

And last but not least, scores can be affected by other parameters too, like uptime, served wallets, and traffic handled by the node.

yegor256 commented 6 years ago

@oltdaniel so basically you are suggesting to add more text to the initial "body" of the score?

yegor256 commented 6 years ago

@oltdaniel what is the objective of all this? What problem are you trying to solve?

ghost commented 6 years ago

@yegor256 Hashing the wallets could change the score from computing power per node to importance of the node since it provides a hashing power with wallets. On this way, a high number of wallets and computing power gives a high score which then represents the importance of the node to the overall network.

yegor256 commented 6 years ago

@oltdaniel how can we validate the number of wallets a node stores? A node can easily fake that number and claim to contain a million wallets. How can we check that?

DronMDF commented 6 years ago

@oltdaniel @yegor256 I have other idea about this.

Current score does not guarantee that the node gives valid information. The node does not pay for the information that gives.

The proposal: Each node mine current content of wallet... <wallet hash> + <nonce> -> <wallet approve value> For each wallet in database.

If client get wallet - node pass approve value to them. Otherwise we do not trust the node.
It is necessary to choose the mining parameters so that the entire base is mined from scratch for quite a long time. Maybe a week (15-20 seconds per wallet is ok). This will increase the entry threshold. This is necessary to complicate the attack 51%. Recently there was a story that some currency was attacked by buying AWS for several hours.
If you need to pay a tax, you can choose any (randomly) node that confirms the contents of the wallet.
Since the node can not immediately confirm all the wallets - You can choose which of them are the most attractive for the site. Pay a tax for an example. Or have to pay. :)