Quote from: rocks on January 06, 2015, 08:02:25 PM

Quote from: justusranvier on January 06, 2015, 07:54:28 PM

Quote from: rocks on January 06, 2015, 07:52:24 PM

(such as including the economic value of a block's transactions in addition to difficulty, this would lower the height/priority of chains that do not include all the transactions that honest miners are including)

That doesn't sounds like a very good idea.

You've just described turning Bitcoin into a proof of stake system.

Gavin had a decent description of it (and other solutions) to address the possibility of 51% attack that withheld transactions a couple years ago, but I couldn't quickly find it.

It's definitely not a POS system though, mining still relies on hashing that anyone can participate in. It would only provide an additional priority advantage to blocks with more transactions in them. If all miners include all transactions in the P2P memory pool then its the same as what we have today. If an attacker decided to withhold transactions, then they'd be at a disadvantage and need progressively more than 51% of total hashing power as they processes less and less transactions.

POS provides rewards proportional to the amount of currency already owned/controlled. The scheme above is not influenced by the amount of currency owned/controlled, so it's definitely not POS.

Are you referring to the IBLT proposal? That would indeed make miners' blocks propagate faster if they include the same set of transactions instead of trying to make their own rules. However, it wouldn't really help in the case of a large cartel, since the cartel's miners would all use the same rules, so it is the others mining (using for example the current largely non-discriminatory rules) that would actually suffer.