RE: A Cursory Overview of How Vaporchains Work
If users can outbid each other to get the block reward you are encouraging centralization. Programmers will be able to scan the blockchain at the last possible moment and outbid everyone else at the very last second, giving them such an incredible advantage that all the less savvy users will be pushed out instantly.
What I'd like to see is a lottery system where everyone who bids has a proportionate chance of winning depending on how much they bid.
Either that or all bids will have to be blind by being cast at the last second on the same block.
I also think it makes more sense to bid with SBD, since it makes more sense to burn our debt... however, I just recently made the claim that SBD should be eliminated altogether so it's hardly worth mentioning.
I also have to wonder how this system stops the bid-winner from posting bad blocks (fake transactions) or posting blank blocks with no transactions.
For the most part I'm very impressed by this project. This is an entirely new consensus algorithm. Proof-of-burn! Whoever nukes the most Steem gets the block reward... crazy. I had no idea there would be block rewards involved.
I have to say, when I heard you were calling it Vaporchain my brain immediately thought Vaporware. The name makes a little more sense now that I know Steem is being vaporized to mine a new chain. Best of luck with proof-of-vapor! Very wow, much exciting.
What could be interesting is to experiment with different ways to distribute the mining rewards. All would be proof-of-burn type, because all amounts are sent to @null in any case, but the distribution mechanism could be different. Here are some:
The mining mechanism could be a parameter in the vapor chain creation command.
Just a suggestion, I'm fine with the current proposal if we can get the first vapor chain faster :)
I also came to that realization when I was drawing this up. Currently, the only deterrent they have is the risk of the bid arriving too late and being invalid. Would you risk sending 100 STEEM in the last 3 seconds in hopes that it'll get added to the last block?
I shall put some thought into that. I haven't considered having more than one miner per block. I will brainstorm on how that changes the game theory and complexity of the protocol.
Something like this could also work, but that would eliminate all the non-techies from participating because they need to scan,verify and broadcast all in 3 seconds. They won't have 10 minutes to do it like I originally planned.
The bid winner can't change transactions because he can't change Steem blocks. All validators detect the same bid winner and just add one VTX ("reward,de_winner,5000,ref") at the end of the Vapor block. It's not something the winner does alone.
Currently, I think the easiest solution would be to keep everything the same and just provide the "bid snipping" software you're referring to. That way nobody has an advantage. And if you want to risk sending your bid too late, you're free to do so :)