[Bitcoin-ml] New Difficulty (DAA) performance

Scott Roberts wordsgalore at gmail.com
Mon Nov 27 12:09:10 UTC 2017


If delays are the only concern, it looks like I was completely wrong
in wanting N=144 a lot lower.  Generally the larger the N, the fewer
the delays (judging by the past 8 days after the initial oscillations
that looked bad).

https://github.com/Bitcoin-UAHF/spec/issues/32#issuecomment-347162795



On Thu, Nov 16, 2017 at 7:31 AM, Scott Roberts <wordsgalore at gmail.com> wrote:
> After the startup period, it looks like the new difficulty algorithm
> will work out OK.  See link below.  This is not an endorsement for
> N=144. N needed to be 50.
>
> Any long delays that result from this algo will not be the result of
> not getting consensus but because the consensus wanted N=144. Even
> Jacob Eliosoff's algo can be messed up by using large N when it needs
> to be a 2.5 hour average instead of a daily or half day averaging.
>
> Hopefully the 5 hours needed to get 6 confirmations was only a initial
> problem caused by the start up period.
>
> https://github.com/Bitcoin-UAHF/spec/issues/32


More information about the bitcoin-ml mailing list