Hey @Frieeze_noot, does it look better with the proposed updates or do you have other feedback ?
If all good I’ll make a recap of the final proposal before posting.
Hey @Frieeze_noot, does it look better with the proposed updates or do you have other feedback ?
If all good I’ll make a recap of the final proposal before posting.
Thanks for the feedbacks and additional informations.
It seems we all agree the goal is to have a balanced team of signers, and based on your datas there’s currently only 2 that are laying behind.
The flaw here is as long as there’s 1 or 2 less efficient signers the rest will be average even if they are not reactive.
As koga mentionned the average time to sign could be a good metric to include (don’t know if it’s easy to get it tho).
If we were to adopt your initial proposal I suggest to limit the rotation to:
If there is too much bad signers/quitters we can reduce the round duration from 6 to 3 months
The data are outdated, will try to update it tomorrow before submitting.
Not sure but at least it’s a great feature suggestion for Den.
However, considering signers are rewarded (not much but still) checking time to sign could incentivize signers to not verify propoerly transactions (a problem already existing in many multisigs)
Considering the discussion yesterday, I still believe my method to calculate is more accurate as it’s based on the real number of transactions executed, but the method of calcul will be a discussion for the next PIP anyway, it shouldn’t block this one.
So if I follow it would be first 3 auto renewed if ok, middle 4 recandidate & bottom 2 are kicked and can’t recandidate for next period.
If max rotation would be 4 per period as suggested above, and assuming 4 middle or some top want to leave, only two can resign during one period if bottom two are kicked.
I’d still like keep the option to kick the 7th if his rate is below 50%, in which case 3 would be kicked. and with the same scenario as mentioned above, only one could resign.
I believe 3 months is better, especially if we’re limiting max kick, as it would enable shorter delay for resignations in worst scenarios but opinions on this are quite different from feedback on this poll: PIP-22: Multisigs rotations framework update - #5 by Dydymoon
Updated Data based on last 249 transactions executed from last rotation to now (previously 204 transactions analysed until January 10th)
No it would be first 5 stay, last 4 needs to re-candidate (or are kicked if sign rate below 40%).
If there are resignations, they should be placed at the end of the list, so we don’t force people to stay.
As long as we rank signers, that’s a risk we have to deal with. But you’re right rating signers performance over time to sign can be too much of an incentive.
Since the initial issue seems to be the time to time/execute, it would be nice to have an average of the whole MS over the period. So we can track the evolution of the efficiency.
I agree with the poll 6 month should be the default rotation period. I only suggested to reduce it if the current team is inefficient enough or if there’s too much bad signer/resignations
Ah yes my bad since we can’t rotate more than 4, then top 5 is auto renewed if agreed.
40% is too low imo, as I was suggesting 50% let’s find a middle ground with below 45%.
Yep ideally we don’t but if we’re kicking 3 signers & 2 wanna resign, only 1 can during this period.
Not sure to understand what you mean or how to get that if you can specify.
Oh ok yes make sense
Alright so if we agree let’s recap the new proposal so we can avoid splitting options & over complicate.
Top 5 signers are auto renewed if they agree.
Middle 6-7 must recandidate, so they could be changed if new candidates are elected
Bottom 2 are auto kicked for at least one period if participation is below 45% (can be up to bottom 3)
Max 4 signers can be rotated at the same time, both reasons included (kicked or resignation)
If two are kicked, up to 2 can resign during the same period. If 3 are kicked, only one can resign.
Rotation period will happen every 6 months but could be exceptionally reduced to 3 months for a period if all resignations can’t be done when required or if there are more than 4 bad signers.
If no additional comment arises, I’ll push the vote tomorrow morning.