Since our difficulty adjustment is different than Myriad we have some more protection for this.
I don't see how that follows. It's really the exact same thing but with max adjustments increased 10x. Historically this has led to insane difficulty swings (even at constant hashrates) which can be difficult to distinguish from actual tomfoolery without going in and comparing timestamps. One slight advantage is that -log(max_diff_down)/log(min_diff_up) is 1.94 for Myriad but only 1.5 for DigiByte, meaning the attacker would need a slightly greater share of the hashpower to attack DigiByte. Since DGB has a higher hash rate anyway, a successful attack seems less likely.
BTW I've updated my pull request on github.
We just saw your pull request! Thank you for contributing it! Long story short, we are vulnerable as well to the same attack if our net hash rate falls too low. But we do, however, have some additional protections built in the Myriad does not.
We are going to issue a hard fork in the next few days. We are going to implement some changes to significantly mitigate this problem. In the mean time our chain is moving along fine, but we can use all the hash we can get. It is very likely who ever was behind the Myriad attack will attempt the same with DigiByte. And it does in fact appear that they have already tried / been trying.
Is that your polite way of saying "Thanks, but we're just gonna merge Myriad's change"?
Please do comment on these "additional protections". As far as I can tell, the difficulty adjustment algorithm was copied verbatim from myriad, and then nMaxAdjustDown and nMaxAdjustUp were increased by a factor of 10.
"Moving along fine" is debatable when every single algorithm's difficulty regularly cycles by a factor of 10. It just gives the illusion of moving along fine because at any given time there's usually at least 1 algorithm going through its low phase.
Jared, almost all questions are addressed, 3 left:
1 Could you please comment on these additional protections (explicit comments and no talking around it)? If they are true, we should do marketing with it. If they are not be transparant and just say there are none. Please see the above conversation
2. Are all our releases open source? (transparant)? Particular digibytetip and ios wallet? If not, why?
3 Currently, an attacker can 51% attack the network with roughly
60% of SHA256D and nothing else. After this change, an attacker with
90% of the SHA256D hashrate and
33% of each of the other 4 algorithms would have insufficient hashpower to mount a 51% attack. Is this true? Source:
https://github.com/digibyte/digibyte/pull/36 So in theory a attacker does not need to have some hashrate in all 5 algorithms (used in marketing of digibyte)? 60% of SHA256D is sufficient?