Quote from: iamnotback on April 14, 2017, 12:26:12 AM

Quote from: IadixDev on April 13, 2017, 07:45:17 AM

Quote from: iamnotback on April 12, 2017, 08:30:37 PM

Quote from: dinofelis on April 12, 2017, 03:12:49 PM

There is a serious inconsistency in how UTXO are referred.
On one hand, there is all the work of having a totally ordered consensus of transactions: the block chain. It would have been extremely simple to refer to a transaction output in a block chain: the block number, the transaction number in the block, and the output number in the transaction uniquely specify the UTXO. No need for a hash, no need for 256 bit !

Seriously you need to stop pretending you know anything about blockchain design.

This is beginners' egregious error.

Lol you just flunked the most fundamental issue of decentralized systems, which is there is no total order.

Well deep down blockchain are still a decentralized database, who preserve total order

Even if the way the chain will be constructed is not ordered, the system make in sort to garantee total order consistent across the network.

Incorrect. Chain reorganizations can happen at any time. PoW is probabilistically (i.e. never) final, not deterministically final.

Thus referencing by specific chains instead of by hash as @dinofelis suggested would be a DDoS security vulnerability at least and other cascading issues.

@IadixDev, that is why you leave the blockchain design work to me. I am expert. You are not.

For the moment im still on the script, it advance good Grin

I will probably give more news on it in a few days.

It should have everything needed to can create a blockchain with custom block validation algorithm, all the parameters to initialize the node for a specific chain with block reward curve, pos, with event handler for p2p messages, and rpc server + block explorer Grin

All this in super simple javascript like language, with the module and event queue.

it could probably be compiled to assembler easily, as it's mostly C call to the framework, and it use only base types that the cpu understand like pointer , integers and simd.

For the other discussion if you follow the thread you will see better where I want to get at, and also why in my idea chain ordering could not matter at all.

But again the main point was not even this, but to show pow is not only about solving consencus etc Grin