Proposal/idea to solve scalability problem

[xavier]

A big problem with bitcoin is the scalability issue. The system is just unscalable right now. The idea and the philosophy behind it are spot on, but I feel like it's still in a `1.0` state. It will need some innovation to survive once the blockchain becomes to big.

So.. I propose a project that embeds the philosophy of satoshi - A philosophy to combine economic incentives with smart technical design.

The problem right now is that each client needs an entire copy of the x GB blockchain, in order to verify transactions. This is unscalable.. as blockchain becomes bigger, bitcoin cannot be used by so many people.

As an alternative, I propose the following.

The blockchain is held only by a small number of centralized nodes. These nodes will be a new part of the bitcoin ecosystem. Right now, there are only 2 types of nodes that are part of the ecosystem: users and miners. In this new system, I propose 3 types of nodes. Users, miners and houses.

Houses just store the blockchain; they are not involved in a competition to produce blocks.

Like miners, the system offers economic incentives for houses to exist. Every time a user makes a transaction, he must pay a verification fee to a house. The house can instantly confirm whether a transaction is valid, by examining its copy of the entire blockchain. A user may want to pay multiple houses for verification (at additional transaction expense) if it is a large transaction, for example.

The verification fee for each transaction will be in addition to existing mining fees, paid to miners who are actually creating new blocks.

Houses can set their own fees; therefore there is a competition for houses to provide the lowest fees possible. There will be a trade off between cost to store the ever expanding blockchain and possible fees from verification of new transactions.

With this new infrastructure, there is no requirement for clients to store the blockchain, and the problem of scalability can be solved.

[1715193961]

1715193961

[1715193961]

1715193961

[1715193961]

1715193961

[1715193961]

1715193961

[1715193961]

1715193961

[1715193961]

1715193961

[molecular]

have you read about SPV clients?

http://bitcoin.stackexchange.com/questions/4649/what-is-an-spv-client

[empoweoqwj]

I'm not sure you understand how it all works at the moment .... miners verify transactions. Most "clients" don't need the blockchain downloaded at all.

[mustyoshi]

When we start reducing the number of full copies of the blockchain, the system becomes vulnerable to attack.

[molecular]

When we start reducing the number of full copies of the blockchain, the system becomes vulnerable to attack.

what kind of attack? deletion of the blockchain? lol.

[xavier]

(1) have you read about SPV clients?

Yes..... but there is no economic incentive there to mine. I am introducing a second transaction fee to provide incentives for people to hold a centralized copy of the blockchain.

(2) "I'm not sure you understand how it all works at the moment .... miners verify transactions. Most "clients" don't need the blockchain downloaded at all."

All clients need the blockchain, unless they are an electrum client

(3) When we start reducing the number of full copies of the blockchain, the system becomes vulnerable to attack.

Not really.... because it becomes harder for individuals to have a copy of the blockchain, so attacking becomes harder.

(4) what kind of attack? deletion of the blockchain? lol.

I think he's talking about the 51% attack.

[DannyHamilton]

There is no design change necessary to implement your system, the current design already supports it.  There just hasn't been anyone who has chosen to charge a fee for providing "house" services yet.  Effectively they are all charging a 0% 0 BTC fee.

As the blockchain continues to grow, in the future perhaps housing the blockchain will become valuable enough that people will be willing to pay a fee for it, and houses will require a fee to be able to continue to offer the service.  For now, many users simply enjoy free use of the "house" services.

[FlappySocks]

I have been thinking about this problem for a while too.  There is no way, someone like Paypal is going to adopt Bitcoin with an unresolved scalability problem.
It's also unrealistic for everyone to keep a copy of the blockchain (mobiles for example). It's bad enough on the desktop, today.

The blockchain either needs to be tamed in some way (mathematically or whatever).

Or we need to give incentives for people to set up large datastores, that can process & verify transactions.  There needs to be lots of them, to a) reduce attack, and b) verify transactions quickly.  My mobile can than pick say 5 at random, and ask them to process my transaction.  Each of which gets a fee.  Fee's will have to go up, which I dont see as a bad thing.  You are after all using someone else's electricity, hardware, and bandwidth.  The data-stores will also provide balances, and transaction history services, so clients dont have to keep a local copy of the blockchain.

[Killdozer]

It almost feels like it's time to make a new child board for the scalability questions/propositions. The search function just seems too difficult for these people...

[justusranvier]

https://bitcointalk.org/index.php?topic=204283.0