Bitcoin Forum

Bitcoin => Development & Technical Discussion => Topic started by: uniqueCoin on December 23, 2017, 04:46:21 PM



Title: 2-blocks shift - could this be a way to have big blocks and keep decentralizatio
Post by: uniqueCoin on December 23, 2017, 04:46:21 PM
as far as i know main problem with big blocks is propagation time, miners who recieve new block earlier get head start in competing for next block, pools can share new blocks faster among their miners further making this gap bigger, hence centralization, so i had this idea, but my knowledge is very basic and i can't seem to find a flaw in it, do you guys think it makes sense?

2-blocks shift:
if creator of block would send his transactions to be added to next after next block
and in his block include transactions passed from 2 blocks before
then after minting new block everyone on the network should already have list of transactions to be added in current block
only hash propagation would be needed to start working on current block
and transaction list for next block would be propagated during working on current block



Title: Re: 2-blocks shift - could this be a way to have big blocks and keep decentralizatio
Post by: Andre_Goldman on December 23, 2017, 05:37:51 PM
as far as i know main problem with big blocks is propagation time, miners who recieve new block earlier get head start in competing for next block, pools can share new blocks faster among their miners further making this gap bigger, hence centralization, so i had this idea, but my knowledge is very basic and i can't seem to find a flaw in it, do you guys think it makes sense?

2-blocks shift:
if creator of block would send his transactions to be added to next after next block
and in his block include transactions passed from 2 blocks before
then after minting new block everyone on the network should already have list of transactions to be added in current block
only hash propagation would be needed to start working on current block
and transaction list for next block would be propagated during working on current block




Are you started talking about the "Pareto Principle"  ..

The 80-20 Rule Explained
https://www.youtube.com/watch?v=F-I-BVqMiNI


Title: Re: 2-blocks shift - could this be a way to have big blocks and keep decentralizatio
Post by: uniqueCoin on December 23, 2017, 06:14:25 PM
no, i'm talking about possibility to increase blocksize and keep decentralization (or even improve it a bit)

what i meant is:

miner who mined new block had added at the bottom sha256 of his transactions list for 2 blocks ahead, he sends out 2 packets:

    1st small packet consisting of 2 things: sha256 of his transactions list and solution for block

    2nd big packet of transactions for 2 blocks ahead

1st packet propagates very fast and since everyone was mining same transactions (passed from 2 blocks before) everyone can easily verify if solution is valid just with this 2 recieved information (they cannot yet verify if sha256 is correct, they trust it because proof of work)

2nd packet propagates slowly, but it has time since we are all working on next block and it propagates during that work (now they can verify sha256 of transactions list and list itself and if something is wrong go back to working on previous block)



i'm just not sure if i'm not making any mistake in reasoning, because my knowledge in matter is very limited
that's why i wanted to ask smarter people if it could work as a solution


Title: Re: 2-blocks shift - could this be a way to have big blocks and keep decentralizatio
Post by: Andre_Goldman on December 23, 2017, 06:36:04 PM
no, i'm talking about possibility to increase blocksize and keep decentralization (or even improve it a bit)

what i meant is:

miner who mined new block had added at the bottom sha256 of his transactions list for 2 blocks ahead, he sends out 2 packets:

    1st small packet consisting of 2 things: sha256 of his transactions list and solution for block

    2nd big packet of transactions for 2 blocks ahead

1st packet propagates very fast and since everyone was mining same transactions (passed from 2 blocks before) everyone can easily verify if solution is valid just with this 2 recieved information (they cannot yet verify if sha256 is correct, they trust it because proof of work)

2nd packet propagates slowly, but it has time since we are all working on next block and it propagates during that work (now they can verify sha256 of transactions list and list itself and if something is wrong go back to working on previous block)



i'm just not sure if i'm not making any mistake in reasoning, because my knowledge in matter is very limited
that's why i wanted to ask smarter people if it could work as a solution

Well I am not sure about data propagation thing .... rule number 1 while pulling new cabes and wires ... if you going to pull one cable at the first time ... why not pull 2 ?

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

 


Title: Re: 2-blocks shift - could this be a way to have big blocks and keep decentralizatio
Post by: uniqueCoin on December 23, 2017, 09:10:46 PM
you can try to upgrade infrastructure

you can try to upgrade code

or you can try to upgrade both


Title: Re: 2-blocks shift - could this be a way to have big blocks and keep decentralizatio
Post by: Anti-Cen on December 24, 2017, 01:47:26 PM
The trouble is not the block size or the timing of the blocks but the sheer volume of data that
needs to be held and processed by each node and the files are already at 200gb and for much of the
time it was only being used by student as they were testing it with small amounts of money.

what might work for a million transactions a day with a few days worth of data will grind
to a holt over the years as the file becomes bigger and bigger without it being archived off and
that's not half the story because a global system needs to cater for a billion or more transactions
a day unless we create Micro-Bitcoin and run it on another platform that uses a tree type data
structure

Like it or not the data won't fit onto a $200 machine and the current design of the infrastructure
is Micky mouse  AND WILL NOT SCALE and yet we need $200 machines (Millions of them) to be part
of the network and will need trusted decentralized coordinators in the system and not just money
crazy miners wasting CPU power doing this PoW bullshit.

Based on public keys we need something like 16 X 16 or better still 32 X 32 top level coordinators
to then organize the $200 nodes into something that can be deemed useful to the network with mining
fees scam becoming a part of the history book since each node will have to earn it's keep using a
system of "Gas" , no gas and the car won't start, do your public duty to earn some.

This network needs to be independent of BTC or ETH and should be seen more like a money motorway
that allows us mere peasants to organize ourselves without interference from big brother or government
and be seen as the core foundations much like ETH is for other alt coins but the connection between
the transport system and the goods (Coins) on it must be broken, Gas cannot be purchased, has no money value
and can only be earned

In other words think about how BitTorrent works and throw in that peers that won't seed will
get cut of or the bandwidth will be reduced down to a trickle of data
 

 

 



Title: Re: 2-blocks shift - could this be a way to have big blocks and keep decentralizatio
Post by: uniqueCoin on December 24, 2017, 05:13:52 PM
blockchain size have very easy fix, just dropping/archiving old parts of it from time to time


For the rest of your post i do believe that gift to humanity would be basic income coin that is Proof of Identity based and inflative, but i'm afraid even if technology is ready, people will never be