And the hits keep coming for the dangerous experiment of maintaining artificially constrained blocks. From 2009-14 Bitcoin functioned with no effective block size limit and mining decentralization post-ASIC was improving. Now, in 2015, that this latent piece of software is allowed to take effect the unintended consequences rear up:
(my bold emphasis below)
The so-called "stress test" was successful. The attack it was testing is underway.
Any miner who cotrols more than one minus 1/phi of the mining power, makes a profit while paying the fees it takes to maintain a permanent backlog of transactions, for as long as the blocks are more than half full of other transactions.
The mining-expenses market is unstable when there are enough legitimate transactions to fill more than half a block, because such a miner will reap more per yuan invested in mining power than any other miner. This leads to a monopoly.
Well, crap. I have done the math now.
I know exactly what's happening. It's a Golden Ratio Attack. This is serious. It ends with a mining monopoly.
What we have now is no longer a test. The test was apparently successful. What we have here is the new business model that was being tested.
Attacks are beneficial for antifragile systems. Without adversity, they never grow stronger.
Perhaps this new Phi Attack will lead to economically balanced tx feeds, as Sunny King has implemented in Peercoin and Primecoin:
Since I mentioned the topic of transaction fees last week, l should mention the exact kind of simplification of fee policy in peercoin/primecoin. As I explained last week, blockchain technology is not well suited as an ultra high volume transaction processing system, thus we want to discourage tiny payments such as a fragment of cent, also known as 'dust' in bitcoin. Bitcoin charges additional fees for these dust transactions. In peercoin and primecoin, each transaction output must have a value at least 0.01. For the next release of peercoin, the output value can also be 0 to accommodate data transactions. But still, dust is not allowed, you either have an output value greater than or equal to 0.01 or a value of 0.
Another simplification is that we don't allow no-fee transactions, so the default 0.01 per KB fee is a requirement, not an option as in bitcoin. This eliminates the complication for user to understand the tradeoff between no-fee and quality of service.
Some might question the consistency of the approach as we move to allow data transaction in peercoin's next release. Data transaction is simply a mechanism to allow user to explicitly declare extra data in the transaction, otherwise user could still stuff data into transaction in obscure ways. In principle transaction fee is the only cost associated with blockchain space usage.
https://bitcointalk.org/index.php?topic=114994.msg11882960#msg11882960