Open theoden-dd opened 7 years ago
Thanks for the code! However, I'm not so happy with the current approach of increasing the broadcast frequency for skipping duplicate nodes. I thought about a different, deterministic algorithm which would rely on a constant cycle frequency:
Optional, to prevent sending known nodes:
The strength of this approach is that we don't need to sample the entire database multiple times to find an unsent node. What do you think?
Excellent idea!
That's generally the same, what I wanted to implement, but achieved by other means.
I'll do that in the 1.2.0 version instead of searching the first unset node after a randomly selected one: theoden-dd#13
The toughest things to find out are:
See comments in the last message for #9.