Quote from: TooDumbForBitcoin on August 19, 2014, 12:03:22 PM

Quote from: fluffypony on August 19, 2014, 11:40:40 AM

Quote from: dga on August 19, 2014, 10:57:21 AM

I'm sure there will be pushback on these as there was to the others, but kudos to him and the Boolberry team for putting it out there for others to read, steal from, and criticize. (Disclaimer: I looked at an earlier draft of this one and provided some minor writing feedback. I'm not an author of it and am not part of the BBR team.)

Busy reading through it - he leads into it with a huge fallacy that is either incredibly naive or very disingenuous of him. When describing CryptoNight he states: "These constraints were supposed to protect hash from GPU and ASIC implementation" [sic]. Literally the first paragraph in the CryptoNote whitepaper that describes the PoW algorithm says: "Our primary goal is to close the gap between CPU (majority) and GPU/FPGA/ASIC (minority) miners. It is appropriate that some users can have a certain advantage over others, but their investments should grow at least linearly with the power. More generally, producing special-purpose devices has to be as less profitable as possible."

Misrepresenting the facts of the matter in a whitepaper, purposely or not, is unconscionable.

Looks like two different ways to say the same thing. "protect from" has the same meaning as "close the gap between" in terms of reducing the ability for GPU/ASIC to skyrocket the hash beyond the capabilities of CPU to do so.

The full sentence in the whitepaper is: "These constraints were supposed to protect hash from GPU and ASIC implementation, but a GPU miner appeared on the scene in 2 weeks after this technology got public attention." Thus, contextually we know that his meaning in the word "protect" is "ensure they do not exist". He considers the very existence of a GPU miner a failure of the algorithm, when, in fact, a GPU miner can and should exist as long as it the performance gap is closed. Currently GPU miners are 2-3x as performant / efficient as CPU miners, and by dga's calculations they shan't exceed ~5x the performance / efficiency. Thus the algorithm has completely succeeded at what it purports to do, and has met its primary goal.