Merge Neuron(s) function

Please see Internet Computer NNS Neurons Can Now Be Merged | by DFINITY | The Internet Computer Review | Apr, 2022 | Medium, which clarifies what will happen with the combined age and dissolve delay.

1 Like

Thanks, is it possible to perform this action in the command line without quill?

The NNS Dapp rolled out this feature today for those with neurons created in the NNS Dapp. If your neuron was created with quill, you will need to merge with quill.

5 Likes

these are seed neurons so not created with quill

Ah, seed neurons. Those are controlled via a private key file you possess, right? This is also a case where quill is necessary, although you have the option migrating to a Ledger Nano instead — once the Nano supports the merge neurons operation.

I’ve absolutely no idea how to use quill, the instructions are a little technical for me

Quick question. Is it possible to merge a neuron created via NNS with a neuron created via Ledger? Eventually, which one has the control after the merge?

Edit: Nevermind, the article provided already answered my question

Very reasonable. I wrote a lot of the docs and I plan to do a new set of revisions to make it 50% easier.

https://wiki.internetcomputer.org/wiki/ICP_staking_options

2 Likes

thanks that would be a massive help!

1 Like

I think you are very right that it is nowhere as easy as it should be yet.

I have spent last few weeks working on simplifying developer experience for deploying smart contracts. I usually jump between user-facing experience and developer facing.

4 Likes

I don’t understand the specific meaning of “If the source neuron has no age, then the final age will be the exact same as just increasing the stake by that same maturity.” Can you elaborate?

The calculation for merging age is:

(source_stake * source_age + target_stake * target_age) / (source_stake + target_stake)

If the source has no age, this simplifies to:

(target_stake * target_age) / (source_stake + target_stake)

This last calculation is identical to what is done for “topping up” or adding stake to an existing neuron, and has the effect of “distributing” the age of the neuron across all of the new stake.

Age is like seasoning on a stake. If you suddenly add more meat, you’ll need to spread the seasoning around to cover the whole thing. The amount of “seasoning per square inch” would be determined using the same type of calculation.

1 Like

@jwiegley
I sent you PM with a neuron merging problem
thanks

comparing to not merged neurons, merged neurons rewards decreased substantially. Merged neurons rewards calculate system has problem or bug, I think.

  1. merged neuron rewards with 4year lockup and 7year lockup are same.
  2. comparing to not merged neurons, merged neurons rewards decreased substantially.

So I thing merge neuron reward calculate system has problem now.

I think the problem you are seeing is a problem with reporting maturity as a percentage of stake. If you have a neuron with 50% maturity and 1 ICP stake, and you merge it into a neuron with 10 ICP, it will only increase the target neuron’s maturity by 5%. No actual maturity is lost, it’s just not reporting the same terms as before because the divisor is different. Does that make sense?

It is not the situation you are saying.

if 50% maturity and 1 ICP stake merge it into a neuron with 10 ICP, 11 ICP 's rewards after merged decrease substantially comparing to 1 ICP and 10 ICP separately.

after merged calculation system has problem of bug, I think.

1 Like

while 1 ICP can get 0.07% rewards and 10 ICP can get reward 0.07% rewards per a day, merged 11 ICP can get 0.01% rewards per a day.

this is current situation.

1 Like

If both neurons were received 0.07%, then the sum should also receive 0.07%. Did you merge into a new neuron that hasn’t voted before? It may take a few days in that case for ballots to be created on behalf of that neuron.

It make sense.

on weekend, there are few chance to vote.
after merged, depending on vote opportunity, rewards could decrease.

1 Like