[PROPOSAL #187][ACCEPTED] V9 Lambda upgrade (with Replicated Security)

Jehan’s note in the post explains the change in terminology.

“ICS v1” requires the full Cosmos Hub validator set (or at least more than 2/3). For each Consumer Chain launched the Cosmos Hub validator set is replicated in full to run that chain. Hence Replicated Security. I don’t know who signed off or what the deliberation process was. While it may be confusing to the public initially Replicated Security is arguably more correct to explain how the system works.

This Informal Systems blog post that compares Replicated Security and Mesh Security is helpful to understand what is meant by “Interchain Security is the name for a family of protocols.”

1 Like

It does not include the Liquid Staking Module. There are several reasons for this.

First, we drastically cut down the release late last year to focus purely on shipping Replicated Security. I think this was the right choice.

Second, the LSM does not actually enable liquid staking. All it does is make it so that delegators can transfer their stake to liquid staking providers without having to wait to unbond their coins, potentially making adoption of liquid staking quicker. Liquid staking works without the LSM.

We will have to see how much community demand there is for the LSM to know how to prioritize it in upcoming releases.


We had been using terminology like “v1, v2, v3” to describe various types of Interchain Security, but this is not a great naming practice and was getting confusing. We are going to use descriptive names from now on, such as “Replicated Security”, “Opt-in Security” and “Mesh Security”.

We honestly should have cleaned up the naming earlier, but we figured it’s better late than never. Sorry for any confusion.


We will have to see how much community demand there is for the LSM to know how to prioritize it in upcoming releases.

Difficult part here is a bit how to measure this. What is “enough demand”?
And is that not a chicken-egg story? Because I can imagine a lot of people hold back on liquid staking their assets, due to the unbonding period and the loss of APR in that time period. So if funds can be moved freely adoption might drastically increase for LSDs.

A few respondents have asked about bugs found during Game of Chains and during the Replicated Security persistent testnet. I’ve compiled a list with links out to issues and commentary on Github, for the more technically inclined.

maybe it would also break the ICS (Interchain Standards) acronym. x)

1 Like

This proposal is a draft and we will be modifying and clarifying it according to feedback received here before putting it on chain.

from a timing perspective, when can we expect the prop to be on-chain? Doesn’t look like there’s much more additional feedback since this draft has been up for nearly 2 months

1 Like

We are just dealing with some release issues. Was supposed to go live tomorrow, might be pushed to Thursday or beyond if we are unlucky.


I still consider the fact that validators cannot opt-out as a major risk for institutional adoption as mentioned in this thread Preparing for Replicated Security - #29 by BuilderBot.

Am I the only one who sees an issue here? Ofc, if institutional adoption is not the goal then please ignore it.

In short, what if a consumer chain token can be classified as a security? Custodial stakers or stakers operating in a similar legal environment might be forced to exit the Cosmos Hub then.

I would say that there is a general opt-out in this situation. If a consumer chain token is deemed a security, and institutional investment feels threatened that they’ll have to exit the system due to receiving a tokenized security as a reward. We as a community have the general ability, and right to vote the chain in question, out of ICS. The Governance gate works both ways.

1 Like

We are on chain with voting period ending 2023-03-07 12:08:29 UTC.


Good Job
Replicated Security

I know it’s been a little while since this topic but I strongly believe that validators refusing to run consumer chains is a much more likely scenario than you’re considering. They hold the votes of any non-voting token holder and if the incentives in proposals don’t line up with expectations (or if expectations of validators aren’t properly managed from the outset, validators won’t want to support. They’ll either vote with their delegated tokens or they could (eventually) consider coordinating off-chain.

What’s wrong with provider and consumer as terminology? I think a consumer is naturally dependent on the provider for things - not just producer/consumer as in “one entity creates and the other consumes” but provider/consumer as in “one entity wants something and the other provides”.

I guess it sounds like the goal of this terminology is to imply a power dynamic in which the provider has all the power and the consumer is forced to do as the provider wishes. But I think we want to make these entities both seem like adults (I feel weird calling Neutron a ‘child’ lol).

And I think it’s in our community’s interest to use terminology that doesn’t imply ‘force’ so much as natural values-alignment. Even if Gaia has the power to disrupt a consumer chain and kick them out, we hope to actually benefit from the relationship and both be motivated to keep it going.

And as a selfish addition here, I feel more comfortable not alluding to slavery, drug use, or kink in my workplace :sweat_smile: call me a sensitive soul? Like I am not bothered by the language in general, but it just introduces a layer of social and emotional complexity that makes my job a lot harder to engage with (especially as a relations/comms person)!

Leader / follower?
Controller / subject?
Manager / contributor?



You’re very right Lexa, I think that all of the terms we might find, eh, there’s a certain domination thing going on.

One of the AADAO members suggested to me that it may be best to just roll with partner as the term, and I think they could be right.