Quote from: Altcoinfanatic on July 10, 2015, 11:05:41 AM

Quote from: HR on July 10, 2015, 10:51:07 AM

I will remain skeptical and adhere to the original analysis and numbers backed by so many extremely experienced and knowledgeable devs. I find it incredible to think that a 51% attact would be possible with only 61% of one algo.

To which extremely experienced and knowledgeable devs are you referring to? 8bitcoder (dev of myriadcoin) and Roshan (dev of Saffroncoin) are extremely experienced and they both acknowledged this problem. I know its hard to believe (i had the same) but one must not be blinded and hide for the truth no?

Read this conversation HR:
https://www.reddit.com/r/myriadcoin/comments/2oj7y5/lowhashrate_51_attack_on_myriad_without_timewarp/

especially this part:
https://www.reddit.com/r/myriadcoin/comments/2oj7y5/lowhashrate_51_attack_on_myriad_without_timewarp/cmnonug
https://www.reddit.com/r/myriadcoin/comments/2oj7y5/lowhashrate_51_attack_on_myriad_without_timewarp/cmnr5gd

TLDR: Myriadcoin and Saffroncoin has patched up their code with some workarounds (but still not optimal). Digibyte has not been patched up yet (but will be in digispeed, we hope).

Thanks for that background material - it's very helpful. You and MentalCollatz certainly are two welcomed additions to the community.

Nevertheless, there's still no statistical analysis there to speak of either. Lot's of great coders, "patching" things without really addressing whether or not the need is real. The probabilistic analysis is severely wanting. How in the world can a "coin [be] vulnerable to 51% attacks by attackers with far less than 51% of the network hashpower"? That just leaves my head in a fog. It must have done the same to 8bitcoder as well since he just left it at that and began working on a "patch" (which is understandable for a coder since the mindset is one of terrible dislike for bugs regardless of their impact).

Okay, and we pick things up from there, no real understanding, at least not documented (so far as I know), of what the real statistical ramifications are, but a 51% attack with less than 51% does cast some doubt, to say the least, on what the real possibilities are.

8bitcoder talks early on about the need to build a "side chain" and that the attacker would eventually catch up to the main network. That is practically statistically impossible, if not completely impossible. Does anyone really think a side chain of equal size to the DGB blockchain could be built? If diff were ZERO then a blockchain of equal size in blocks could be theoretically achieved, but with a zero diff, the "work value" would be infinitely inferior. In order to build a side chain of equal value, we'd need Superman to reverse time for us so that we could start at the same point in time. And don't come back at this by saying a forked side chain would work since by the time the attacker were to catch up that "fork" would be long known to be just that.

Of course, everything I'm saying is just off the top of my head, but without any serious statistical analysis on the part of those that say this is a reality - a 51% attack with less than 51%, or even with 61%, as has been said here - what I'm saying holds the same weight (if not more since common sense tells us that less than half is not equal to half).

Common sense tells us that a 51% attack requires a 51% control of the ENTIRE network, not 51% of 20% of the network, unless you're able to stop the other 4 algos from discovering new blocks over a short period of time, something which no-one has ever said is possible (will this be the next shoe to drop?) and something that doesn't happen regardless of how much one individual algo's hashrate spikes.

WarpTimer starts that thread by saying that in "theory it could be carried out on a single CPU." Seriously? What about the diff needed to generate the amount of work necessary in order to be equal to a real SHA256 block generated on the real blockchain? With a single CPU? Come on, where are we drawing our assumptions from, and, more importantly, just how are we validating them?

He even goes on to say that there needs to be a spike in SHA256 diff in order to begin the attack, and the attacker is going to be the block finder with one CPU? Cheesy

Let alone try to mine 2 other algos?

As always with the 51% attack, we're back to the need for some serious computational power, and, in this case, the failure of the other algos.

Don't know how to leave you on this other than to say that on this level I don't consider reddit and BCT comments to be anything more than that, comments, and I certainly don't put them into White Paper category, much less documented research.

BTW, do you have a link to where 8bitcoder moved the discussion on github?