Topic: Bitcoin becomes a transport protocol (?) (Read 1489 times)

It is described in the link in my sig. Initial coin distribution is trickier, but could be solved by using a bitcoin-like chain until a sufficient amount of shares are purchased. Or by making it a community effort and awarding shares prior to the start of the chain.

Of course, I'm just saying the problem that Bitcoin solves in a novel way with mining is the very specific double-spend problem, everything else it does doesn't need a blockchain. (well, maybe initial coin distribution too)


If you can figure out how to do that and be sybill-attack resistant you can be the next Satoshi.



Expanding on my last message, actually, come to think of it, we don't need a one-to-one relationship... and maybe spam isn't as big of an issue.

Lets suppose we want to start a pure alt-coin. Lets assume the initial coin distribution problem is solved, and multiple participants have coins they want to create transactions for. Like Bitcoin they can broadcast their transactions on a flood-fill P2P network, letting everyone know about them.

Anyone is then allowed to build a block consisting of a merkle tree of transactions, hashed with a header that includes a reference to a parent block. They broadcast this block on the alt-chain P2P network, and simultaneously broadcast a specially marked Bitcoin transaction including the hash of the block on the bitcoin P2P network. Dos attacks can be avoided by dropping nodes that transmit blocks without submitted the associated Bitcoin transactions.

Unlike Bitcoin the rule is that multiple parallel blocks, that is blocks with the same parent block, are allowed. Transaction validity however is defined by being in the block whose marked transaction in the bitcoin block chain is first. Transactions that are found to be invalid can be ignored simply ignored, and blocks with excessively large numbers of invalid transactions can be ignored in their entirety. You'll may also still want some sort of payment-based rule to discourage creating huge numbers of blocks, such as a mandatory fee, maybe tied to the average transaction fee on the bitcoin network. You can determine if your peers are hiding transactions from you by just scanning the bitcoin blockchain, although come to think of it this mechanism is susceptible to griefing by creating bogus marked transactions...

This does do a decent job at minimizing the extra junk left in the Bitcoin blockchain though by allowing transaction combining. It's also possible to "merge-mine" multiple such alt-chains with just one transaction.


Oh, and here's another idea: going back to the "highest-paying block wins" it'd actually be possible to solicit donations to pay the fee. The proposer of a block would create a transaction template and solicit signatures from the donators. The interesting thing here is donators can donate to multiple candidate blocks that they are happy with, and only one block would ever get accepted as the others would look like double spends. Griefing is of course possible, just like similar coin-mixing proposals, but solutions to the latter may be solutions to the former. Again, I dunno if this is actually useful, but maybe someone will come up with an application for the idea.

Or you could find a decentralized entity that determines who puts what transactions in the chain, without mining.

You can't trust an entity that is held accountable by no one.

Also remeber that the content of the blockchain is chosen in a way that's effectively voting, with the vote being calculated proportional to the computer power possessed by the people mining new blocks. Essentially the vote of miners decides what transactions go in the block chain, and very importantly, in what order.

If you could find a trusted centralized entity to determine what transactions, and in what order, should go in the blockchain, you wouldn't need mining at all.


As a thought experiment, you could make an alt-blockchain using the bitcoin blockchain where the order of blocks and transactions was determined by who was willing to spend the most bitcoin to insert a block. The way this could work is like in bitcoin you would have some mechanism to broadcast transactions to everyone in the network, such as a peer-to-peer flood-fill mechanism. Also like Bitcoin someone who wanted to generate a block would collect all the transactions and tie them together in a merkle tree, which when hashed with the block header leaves you with a single cryptographic hash.

Now to propose a blocks you would create a transaction spending bitcoins in a specific way, for instance by giving them back to miners in transaction fees, or by sending them to an unspendable address. Included in this transaction would be your hash of the block, and some special marker so that people can scan the bitcoin blockchain and be assured of finding every such special transaction.

Whoever's transaction spend the most bitcoins wins, similar to how in bitcoin a block wins when it has the most difficult proof-of-work solution. Essentially you're voting not with computer power, but money. This might be a reasonable design decision for specific applications.

Interestingly this system actually has better protection than bitcoin itself from sybill attacks: since every candidate block has been specially marked you can simply scan through the bitcoin blockchain and find every submission. Even if every peer you are connected too lies about the state of the alt-blockchain, you can at least tell their lying because you have proof that an alt-block has been submitted with a higher payment value.


On the other hand alt-chains that don't have a cost associated with proposing a block will find themselves spammed. Lets suppose the transaction in the above example just had to be marked; a winning payment amount wasn't required. While you could still decide on a block order by following the order transactions were seen on the bitcoin block chain, there isn't anything stopping someone from submitting lots and lots of blocks with little or no transactions in them, creating an incentive to spam both Bitcoin and the alt-chain.


Of course systems with marked coins don't have this problem, as there is a one-to-one relationship between alt-coin transactions and bitcoin transactions.

A bitcoin is a crypto-signed message.

The blockchain is an archive of signed messages.

I guess we need price discovery to find out how much each additional merged chain will cost to have it added to the merged mining repertoire of the typical miner or of 51% or more of mining hash power that is out there.

Quite a few chains are sitting in the wings waiting for some reasonable assurance that they will have enough mining if they come out into the open.

Hmm I wonder if assurance contracts can help them with that?

-MarkM-

This could also be an anchor point for competing crypto currencies (e.g. such as PPcoin): Cheaper services can build on top of cheaper crypto currencies. That is in particular interesting, because the crypto currency only has to have the following features:
- secure network (double spend prevention, constant generation of blocks).
- prevent counterfeiting.

The minting of the underlying cryptocurrency is then ONLY important for securing the network and is less a function of the scarcity of the moneysupply, because things like limited moneysupply can be build on top of the actual crypto currency.

With some of the recent developments I see across the forum, I begin to wonder what the fate of bitcoin holds. It seems to be that people start putting layers on top of the bitcoin transaction protocol to achieve additional features and functionalities.

To name a few:
- Colored BTC: https://bitcointalksearch.org/topic/chromawallet-colored-coins-issue-and-trade-private-currenciesstocksbonds-106373
- Smart Property: https://bitcointalksearch.org/topic/smart-property-41550
- Assurance Contracts: https://bitcointalksearch.org/topic/m.1201508

It looks to me that cryptocurrency is the http of the world wide web. However, unlike http, traffic in the bitcoin network comes at a cost. The reason is that proof-able and irreversible transactions cost money (and energy).

Thus the more services build on top of bitcoin, the more crowded the network becomes and ultimately the cost rises. It will be interesting to see how that equilibrates, that is : what services can "afford" to use bitcoin for irreversible transactions?

Please list here the promising new services which use bitcoin as a transport layer if you know any.