Quote from: thezerg on August 13, 2015, 06:28:23 PM

Quote from: Erdogan on August 13, 2015, 04:22:57 PM

Quote from: Peter R on August 13, 2015, 03:17:41 PM

Quote from: Zangelbert Bingledack on August 13, 2015, 03:02:26 PM

At this point I'd say just find a way to put the forks on the market and let's arbitrage it out. I will submit if a fork cannot gain the market cap advantage, and I suspect the small-blockers will likewise if Core loses it. Money talks.

I had a strange idea recently: what if we don't even bother with BIP100, BIP101, etc., or trying to come to "consensus" in some formal way. What if, instead, we just make it very easy for node operators to adjust their block size limit. Imagine a drop down menu where you can select "1 MB, 2 MB, 4 MB, 8 MB, ." What would happen?

Personally, I'd just select some big block size limit, like 32 MB. This way, I'd be guaranteed to follow the longest proof of work chain, regardless of what the effective block size limit becomes. I'd expect many people to do the same thing. Eventually, it becomes obvious that the economic majority is supporting a larger limit, and a brave miner publishes a block that is 1.1 MB is size. We all witness that indeed that block got included into the longest proof of work chain, and then suddenly all miners are confident producing 1.1 MB blocks. Thus, the effective block size limit slowly creeps upwards, as this process is repeated over and over as demand for block space grows.

TL/DR: maybe we don't need a strict definition for the max block size limit.

That is exactly what I think. The miners will have to try it out or get some feel of what they can do through other channels (social media, conferences, node versions), including associate with other miners. As long as the association is voluntary, it will not form a monopoly.

yes, this has been considered and discussed before. The danger is that a large block miner cartel might develop naturally whose blocks put small-bandwidth players at a disadvantage. But as others have mentioned, some people are at an electricity cost disadvantage, some bandwidth, some something else... basically it would just be another metric to take into account as you site your miners.

So I would be 100% for this if miners could only work with real txns. But a miner could fill up a huge block with a bunch of "fake" (unrelayed, fee pays to himself) txns to artificially drive up network costs. Its too bad Bitcoin doesn't have the "pay portion of fees to miner pool, receive portion for the next N blocks" feature... that idea closes a lot of miner loopholes.

But regardless I'm not sure if this "loophole" really is one; it does require 51% of the network to be as connected as you are and willing to process your monster garbage block. I have a hard time believing that miners would do so since over the long term they need bitcoin to succeed. More likely (as you guys suggest) they'll just configure their nodes to ignore monster blocks unless > N deep in the chain.

First I am not afraid of cartels in the free market (laws against market actor collusion is rather a statist notion of fear of markets, in reality it is the regulator who is the monopolist intruding with undue force and destroying the market), and second, I don't think a miner will act to waste bandwith, if that were the case, we would already have seen it, since unrestrained block size has been the situation for the first 6 years.