Closed gitferry closed 4 months ago
I don't see any concerns on liveness or impact on reward distribution from my side. Do you have any concerns?
Per yesterday's discussion. maybe we can worry about this later?
Per yesterday's discussion. maybe we can worry about this later?
Yep, will probably compose a RFC for this. Merging this PR now
This PR is part of #682 and a follow-up to https://github.com/babylonchain/babylon/pull/681#discussion_r1658290121. Specifically, this PR:
LivenessDelay
tox/finality
params, which sepcifies a number of heights after which we examine the liveness of finality providers at a certain height. For example, the current height is100
andparams.LivenessDelay = 3
, then we will deal with liveness issues at height97
. This is introduced due to the fact that we allow finality sigs after quorum.inactive
flag of a finality provider according to the same algorithm that setting theinactive
flag totrue
. For example, in a signed window of100
, we determine a fp to beinactive
if it misses more than 50 blocks. So, if this fp is alreadyinactive
, it will be reverted at the height where the missed count is reduced to50
in the window.