Baruk Khazad!
Woodcutters, lend me your chainsaws!
I come to not to bury the hatchet, but to raise it.
Intro:
After a brief respite from the continual hash withdrawl attacks we have seen on the network since November, we are back to the same pattern.
A relatively large GPU farm moves in every time the difficulty drops, and moves out once it goes up.
The activity is clearly not profitable for the attacker, but that is neither here nor there - we have to deal with it.
Effectively the network becomes several times slower to confirm transactions, and more sporadic. Otherwise, business as usual.
Announcement:
Being a public coin, how you deal with this is of course up to you. I plan to take advantage of the ongoing attack by leveraging it as a stress test.
So, expect a bunch of spam TX in the upcoming week. There will also be more fees collected by woodcutters. So, now would be a good time to turn
those spare GPUs or other skein cycles onto the network.
Lets see what we can do.
Enjoy!
I'm not sure if this is something you thought about but having a secondary mining algorithm that can keep the blocks going when the main chain is stuck in a high difficulty retarget works.
Ever since Scrypt ASICs arrived a bunch of Scrypt-only coins are suffering from the same issue; ASIC miner starts mining with enormous hashrate which causes the diff to get kicked to the moon and once the big hasher leaves for greener pastures the blockchain grinds to a halt until some small miners eventually get lucky to find blocks and slowly reverse the retarget mechanism.
So in essence, a single difficulty retarget always relies on - one way or another - the single blockshain's block finding frequency which can be "attacked". The reason for that because difficulty retarget mechanisms can only ever change the difficulty once a block has been found. But if the diff is super high, it will stay that high until the next blocks. And for difficulty retargeting methods that never kick the difficulty "too high" only means the ASIC miner will mine way more blocks than it was intended.
Having multiple algorithms (PoW or PoS) eliminate that.
I'm not sure about block withholding attacks but I'd imagine having multiple algorithms would also cripple their efficiency if not eliminate them.
