Topic: An idea for a Bitcoin scaling alternative to LN (Read 922 times)

What i did in syscoin was burn or deflate fees for syscoin services under 5 per block and inflate above that such that yearly inflation maxes out at a few pct annum.. thus setting up a gdp type metric for supply evaluation in real time as syscoin services qualify as real work being done. Nash would qualify it as ideal money.

Finally had a constructive criticism on Bitshares forum:


Yea, I thought about this problem already.  What I posted was a preliminary idea on how to get a system like this to work.  The base layer tokens would just follow the longest chain rule and probably need to have an expiration date after a certain number of blocks deep.  A reverse confirmation system if you will.

r0ach

Post this on steem if you want to get Dan's attention.  The three of you could have quite an interesting conversation for the rest of us to ponder...

One difference between your proposal and my design appears to be that in yours the second tier nodes are static (or replaced periodically) similar to DPoS. Yours I think has some vulnerability around the selection of those nodes, and I'll need to analyze that when I have some free time.

In my design, any one can standup a 2nd tier node at any time. My design appears to retain Satoshi's open entropy concept at all levels. I am quite proud of my design. Eager to get it launched so I can publish the white paper and we can begin to peer review it.

Also of course mine will be the first attempt at unprofitable PoW, and I don't think you are incorporating that aspect, which I believe is absolutely essential to avoid a centralized clusterfuck.

Well, I believe proof of burn is the most powerful underutilized tool in cryptocurrency.  If you use proof of burn with the actual currency itself, you usually destroy the economic system in some manner.  You either have to make the money supply more inflationary than what is burned just to be safe so the supply doesn't go to nothing, or implement an actual central bank in the coin that monitors the supply on the fly and alters it.  Both of those options are not good in my opinion.  However, if you split the protocol into two different coins, each with a different purpose, you can then burn the one that isn't used as a currency and solve some problems without affecting the economic system.

I've shown this system to a lot of people so far and nobody has been able to find any drawbacks to it over regular Bitcoin.  McElrath claimed it was prone to DOS attack, but I fail to see how this system has a larger attack surface than Bitcoin.  If you had 1001 block validators for instance and they're locked in for a 1 month time period, but the landscape of block validators changes a lot each month since you have to burn the mined coins to become one, how could that possibly be a bigger attack surface than Bitcoin where the amount of pools is very small, everyone knows who they are, and no coins are burned to become one so they change less.

I'm not trying to create the perfect system here, just one that improves on Bitcoin without having any serious drawbacks compared to the current one.  If you get much higher TPS and all the claimed negatives like McElrath's example are trivial at best, then that would be a victory to me.

That aspect sounds like an opening description for mine, but then after that you seem to diverge from my design.

I haven't yet fully evaluated your idea, because my mind is off on Streemit's design issues at the moment.

I rewrote the original post since it seemed to be confusing for some people.  Should be easier to figure out now.

My design could not be merge-mined with Bitcoin due to the radical difference in the block design, so it wouldn't gain Bitcoin's mining security as a side-chain. I think possibly there can only be one winner-take-all of my design because the PoW is effected by the payers so one chain will have the most security. It is like Bitcoin dejavu 2009 again, perhaps...

A lot depends on timing and LN, etc.. So I need to not talk and just release something.

I just cleaned up the text a bit to have it make more sense (probably after you already read it).  My example could be implemented as a Bitcoin sidechain or as a replacement to Bitcoin itself by everyone just proof of burning into it.  I was under the impression you were making a for-profit system, since as you said, you're not a communist and want to get paid to work.  What exactly prevents Bitcoin from implementing your system for free?  Since as you stated, it's probably similar or identical to what I already posted.

Or, let's say what I posted is not similar.  Is it even possible for your system to have value if Bitcoin can implement it after the fact of you releasing?  You can't really beat the network effect if it's possible for Bitcoin to proof of burn or fork into it.

r0ach I can't comment, because you may or may not be getting closer to my design. So I have a conflict-of-interest. I will recuse myself until after I launch.

There are two main ideas someone has to keep in mind when reading this:

1)  Mining is inherently just purchasing a futures contract

2)  The similarity of the security model of Bitcoin NG in relation to this one

I came up with this idea while trying to figure out if there was a simpler way of creating a second tier scaling system for Bitcoin like Lightning Network, except more closely integrated instead of being completely detached, while also not fundamentally altering the current consensus system.  The first thing that has to be done is change from a single coin protocol to one that utilizes two different coin units, each for different purposes.

The regular 21 million Bitcoin supply without inflation is untouched, but it rests on top as a second tier layer.  You have to proof of burn your Bitcoins into this system, you can't directly mine it.  The first tier layer that you do mine, "nodecoins", is mined solely to determine who becomes the block validators (think pools).  As you will see, separating the protocol into two different coins has huge benefits, one being that you can aggressively use proof of burn on the coin that isn't used as money and reap the rewards of that highly powerful but underutilized tool without affecting the actual economic system.

The protocol would use a fixed number of deterministic block validators like DPoS for it's 2nd tier layer (101, 501, 1001 or other number), and every 30 days the people who mine the first tier, base layer nodecoins would have to bid the coins they mine against each other in order to become one of these block validating nodes.  This is accomplished by each nodecoin miner locking their bids in a CLTV-like mechanism and every 4320 blocks (30 days), the top bids are burned and they become the block validators of the system for a 30 day time period.  

As you can see, it's a normal, open loop PoW system with a DPoS-like, deterministic system built on top of it.  Since the deterministic system is the second tier layer, it inherits the security model of the first tier layer, making it not a real fundamental change of how Bitcoin already works.

What is the purpose of becoming a block validator?  Each winning bidder whose mined coins are burned gets a share of the transaction fees (this is one part I'm not sure how to solve besides doing a flat rate per transaction).  The act of mining in order to proof of burn coins to become one of these block validators is essentially doing futures market speculation that the amount of transaction fees you receive will be worth more than how much you spent to do the mining.  It's very similar, in a Bitcoin-NG type of way, to the Bitcoin system that already exists.  It's just the second tier network is built into the protocol itself to expand capacity instead of being separate like LN.

One of the things that people claim prevents miners from acting up currently is the sunk cost on their equipment and not wanting to make it's value useless.  This same functionality can easily be mirrored in this system by just time locking the transaction fees they receive to not be spendable until 40 days later.  That means the transaction fees you receive on day one of becoming a block validator will become spendable 10 days after you no longer are one.

The fact that the nodecoins you have to bid to become block validators are burned, means it's also harder to monopolize block validation in this system.  If a big nodecoin miner wins two bids in a row, all of those coins are burned and the smaller miners catch up to him.