[Bitcoin-development] Proposed alternatives to the 20MB step function

[Thomas Voegtlin]

Le 28/05/2015 17:53, Gavin Andresen a écrit :
> 
> So my straw-man proposal would be:  max size 2x average size over last 144
> blocks, calculated at every block.
> 

I like that idea.

Average is a better choice than median. The median is not well defined
on discrete sets, as shown in your example, and there is no need to be
robust to outliers, thanks to the max size.