Interchain Security and the Game of Chains

There's rarely a dull moment in crypto, but the industry is in an especially weird place right now.

FTX's rapid blow-up, and its far-reaching consequences, have sent projects across the space spiraling. And this is, of course, just the latest in a series of developments this year that have shaken the crypto ecosystem. How'd we get here? Researcher Molly White and Bloomberg's Matt Levine have some helpful explainers:

Molly White:

  1. The FTX collapse: the latest revelations (part three)
  2. The charges against Sam Bankman-Fried and the first FTX Congressional hearing

Matt Levine (paywalled, free with newsletter sign-up):

  1. FTX's Balance Sheet Was Bad
  2. How to Do Fraud at a Futures Exchange

We're watching along with everyone else to see just how this complicated series of dominoes will continue to fall, and our thoughts are with the many people affected.

Yet, for as much as the scandals and scams rampant in crypto may dominate the conversation at the moment, there are more hopeful developments taking place in some of the field's less visible corners. In this piece, we're going to share an update from one of them: the Cosmos, a unique crypto ecosystem on the cusp of a game-changing innovation.

Let's dive in.

The Cosmos and Its Hub

Historically, when two different blockchains have wanted to share data with each other, they've required something called a "bridge," an application that allows objects on one chain to be represented as "equivalent" objects on the other. Each side of the bridge retains distinct infrastructure—separate code, separate hardware, etc.—but these two sides can now interact, in a way, "on-chain." Bridges have succeeded at enabling formal coordination across networks, but have proven over time to be rather problematic. This year alone has seen a number of damaging, high-profile attacks made possible by exploits in bridge code.

Bridges aren't the only way to share data among blockchains, though. Modern chains, unburdened by the technical constraints of longer-running networks like Bitcoin and Ethereum, are being built from the ground-up with modularity and interoperability in mind.

"The Cosmos" is an ecosystem of such chains, united by their adoption of a set of standards called the Inter-Blockchain Communication (IBC) protocol, which allows compatible networks to share data natively among themselves; no bridges required. The basic vision behind the Cosmos is for the Web to be run on a collection (a… cosmos?) of "app chains," independent blockchains that serve their own purposes while connecting seamlessly with other networks. Sitting at the heart of this vision is the oldest of these app chains: the Cosmos Hub.

In function, the Cosmos Hub was built to be simple. Like Bitcoin, it eschews complex smart contracts in favor of one primary use case: transferring tokens (in this case, ATOMs, the native tokens of the chain) between accounts. In concept, however, the Cosmos Hub serves a more complicated purpose. As the first, and most battle-tested, IBC network, the Cosmos Hub is the de facto financial heart of the Cosmos, with ATOM acting as a kind of reserve currency for the broader Cosmos ecosystem. The Interchain Foundation, one of the core development outfits behind IBC and the Cosmos, described this vision for the Cosmos Hub as a "port city."

Since the Cosmos Hub went live in 2019, we've seen this vision largely play out, with the ATOM serving as the most liquid on-/off-ramp for capital to flow through the Cosmos ecosystem. But, while this simplicity has remained a virtue for many in the Cosmos, a growing segment of the community has pushed to expand the Cosmos Hub to do more.

In April of this year, that “do more” segment nearly succeeded, with the introduction of a governance proposal to enable new smart contract functionality on the network. That particular initiative failed, but it set the stage for an even bolder proposal that was made public at the Cosmos's premier event, Cosmoverse, in September. This new initiative, branded "ATOM 2.0," called for a set of sweeping changes that would introduce new features to the Cosmos Hub and, in theory, drive more value to the ATOM and its holders. After weeks of discussion, counter proposals, and counter-counter proposals, ATOM 2.0 ultimately failed to pass a community vote as well.

So, where does that leave the Cosmos Hub today? With its long-term direction uncertain, the Hub's community has turned its attention to the launch of a less controversial, and long-simmering, feature for the network: Interchain Security (ICS).

Interchain Security, Briefly

We noted above that even if two blockchains are connected by a bridge, they remain two distinct networks, with separate software, separate hardware, etc. The Cosmos isn’t so different, in that, even if two blockchains are IBC-compatible, they remain separate chains with distinct back ends. They may be natively interoperable with other networks in the Cosmos, but they're still responsible for maintaining their own infrastructure.

This presents a number of challenges for new chains, including the particularly vexing problem of spinning up their own healthy, well-functioning validator sets. The Cosmos Hub, to take an example, runs on a set of 175 active validators, while Regen Network, to take another, has its own, distinct set of 75 active validators. Each network within the Cosmos is responsible for maintaining its own respective validator set. In an evironment where validator operators are short on resources and different chains are competing for their attention, this can be a pretty tall order.

ICS solves this problem in a clever way, by simply allowing new chains to run using the validator set of an existing chain. In the official ICS vocabulary, the new network becomes a "consumer chain," because it's "consuming" (or, "leasing") security from the existing network, which in turn becomes a "provider chain" because it's "providing" security to the new network.

This represents a complicated bargain for the Cosmos: on one hand, ICS will make it easier than ever for new app chains to launch with access to healthy, robust, and field-tested validator sets (good for decentralization!). On the other hand, it explicitly affords even more power to entrenched validator sets, scaling their centralization characteristics from one chain to, potentially, many others (bad for decentralization!).

No matter how you slice it, ICS will make the health of validator sets in the Cosmos more consequential than ever. It’s critical, then, that validator operators are well-prepared for its launch, and, this month, we joined dozens of other infrastructure teams in completing the final major prep program ahead of ICS’s 2023 rollout. It was called the “Game of Chains.”

On Court in the Game of Chains

Game of Chains was what's known as a "testnet program," which is like a series of exercises in a dummy version of a network, one that involves only "fake" transactions and valueless tokens. Safely removed from real-world consequences, testnet programs act as laboratories for teams to explore (and, often times, fumble) ideas before taking them live with users.

Events like this are typically invisible to non-DevOps engineers (that is, most people), so we wanted to share a closer look at what Game of Chains was, and how Chainflow participated as part of our longstanding history of supporting the infrastructure behind the Cosmos.

Game of Chains was organized and coordinated by Hypha Coop, a cooperatively-owned consulting team with a growing footprint in the Cosmos ecosystem. You may recognize them as the custodians managing the Cosmos Hub's ongoing community governance process via the @CosmosGov handle and the Cosmos Forum. In fact, we worked with Hypha earlier this year to launch the Validator Profiles section of the Forum. Joining Hypha in organizing Game of Chains were the Interchain Foundation and Informal Systems, two core developers on IBC and the Cosmos.

The program was structured as a dummy environment, and a series of tasks for us and the other participants to complete within it. Essentially, these tasks served as practice reps for all the major infrastructure-side mechanics that ICS will call for in the real world. We tackled them in three phases:

  • Phase One: the organizers launched an initial provider chain, then we voted to bring two new consumer chains online, and spin up validators to run them.
  • Phase Two: we voted to bring even more consumer chains online and run them.
  • Phase Three: all bets were off, and participants could create our own governance proposals and launch our own chains.

In all, we trialed spinning up and shutting down dozens of new chains, each interacting with ICS features in different ways.

To get a sense of what, exactly, that looked like, check out the testnet governance page, where you'll find the nearly 100 proposals processed over the course the event, each calling for approval from the community of testnet participants to do things like launch new chains, shut down old chains, and modify various parameters across this interchain mini-ecosystem. Put simply: Game of Chains was a ton of work for everyone involved.

You might be wondering, then, "Why participate in the testnet at all, if it’s not real, it's hard, and you can just support the mainnet later?" Well, first, and most importantly: we're an infrastructure team; this is what we do. We've been supporting testnets and other infrastucture efforts in the Cosmos for more than five years, and it remains a core focus of ours to build the infrastructure powering the ecosystem. Testnets like this are crucial for making sure that that infrastructure is ready for primetime. Second, and perhaps less visibly: Game of Chains, like some other "incentivized testnets," was designed with a built-in reward structure.

You'll notice, in the official Game of Chains task list, a section detailing the "point system." Any time a team completed a task, that team was entitled to official "participation points." Throughout the program, we and the other participants submitted evidence of the various tasks we finished, and our collective participation was tracked on a shared "leader board."

The deal for testnet participants was essentially this: the more points you earned from completing tasks, the more rewards (like, say, ATOM delegations to your mainnet Cosmos Hub validator) you could be entitled to in the real world, later.

So, how’d we do? If you're feeling really frisky, you can find our validator—along with each transaction, vote, etc. that we participated in—here. Otherwise, here's a high-level breakdown of everything we did to support the testnet:

  • Phase One: we voted to launch the first batch of two pre-defined consumers chains, and we launched validators to begin creating blocks for these new chains.
  • Phase Two: we voted to launch the next batch of, this time, four pre-defined consumer chains; and we launched validators to begin creating blocks for them.
  • Phase Three: we voted to reject invalid proposals; we voted to accept new consumer chains; we voted, in fact, on 70 of the 87 available of proposals; and we started to launch a type of support node called a "Relayer," but were cut short when the test chains were halted earlier than originally scheduled.

As of writing, the leaderboad was still being updated to reflect our final point toal. We look forward to seeing our final point score at the end of the process later this month.

Regardless of our final score, we're coming out of the program feeling inspired. Having spent years operating on, and helping to design, testnets, we can say with some authority that the Game of Chains testnet was one of the most exciting testnets we've ever joined. A few reasons why:

  • ICS will be a game-changer for the Cosmos: We've been supporting the Cosmos since the very beginning, and this represents the biggest innovation since the launch of IBC. We're eager to see how this new feature set translates to the real world, and the tesnet organizers designed smart program tasks that helped to battle-test the newest functionality. We're feeling ready.
  • Hypha Coop delivered a well-organized experience: Testnets are chaotic, and often deviate from the planned roadmap in surprising ways. Having Hypha, with their extensive familiarity with the Cosmos community, on board helped to provide a comparatively smooth and productive program.
  • Our fellow participants were dialed-in: Something that most people don't have visibility into is that a lot of what makes crypto tick is a bunch of Discords and group chats among infrastructure operators, all helping one another troubleshoot various issues. This is true of mainnets, and doubly true of testnets, where we're chatting constantly with other teams to break down tasks and provide technical support on tricky infrastructure issues. Game of Chains was no exception, and we're thankful for all the support we received from other participants. We feel it’s especially important to note here that the most active contributors to Game of Chains were the teams behind smaller, independent validators. Just like on mainnets, bigger validators and exchanges stayed, for the most part, on the sidelines. Something to keep in mind when considering where to delegate your tokens: it’s the indie teams who make crypto work!

As we turn our attention to the mainnet launch of ICS in early 2023, we're feeling sincerely grateful for all the hard work teams across the ecosystem are putting in to make sure it's a success. But we're also feeling a bit nostalgic coming off this latest testnet. You see, Chainflow was born from the earliest efforts to build out Cosmos infrastructure, and, as we look to the future of the ecosystem, we've also been reflecting on our history growing up in the Cosmos.

Growing up in the Cosmos

ICS represents a massive undertaking for the Cosmos community, and a significant innovation in launching new blockchains. But this is just the latest in a long line of innovation from the Cosmos ecosystem.

Take, for example, Proof-of-Stake consensus. While Ethereum transitioned only a few months ago to a Proof-of-Stake mechanism, Cosmos blockchains have been trialing Proof-of-Stake for years. Back in 2017, the early Cosmos Hub team convened a working group of infrastructure providers to explore how to operate a sustainable Proof-of-Stake network. This set the stage for a number of blockchain infrastructure projects that have grown significantly in the years since—Chainflow included. Our Founder, Chris Remus, launched the organization on the back of participating in this original working group.

In 2019, the Cosmos held a pioneering Proof-of-Stake testnet program called the "Game of Stakes” (the prequel, in a sense, to Game of Chains). Alongside the other participants, we battle-tested many of the ideas, like staking incentives and validator rewards, that have now spread across the Proof-of-Stake ecosystem to networks like Ethereum, Solana, and others. You can see some of our early feedback on the experience, here.

Game of Stakes was one of our most formative experiences as an enterprise-grade validator operator. Just as we’re earning rewards from the Game of Chains program, we were a Game of Stakes “winner," and, as a reward, we earned some initial grant tokens that helped us kickstart our operations. Our Cosmos Hub validator was our first mainnet launch, and remains our longest running validator in a portfolio that now includes 16 (and counting!) mainnets.

As an active participant on the backend of the program, we got to work with an incredible community developers and infrastructure builders who shared our values around making the digital economy more accessible and inclusive. And, perhaps most importantly, our experiences in these early years of the Cosmos showed where Chainflow could help to contribute to the broader movement for decentralizing the Web.

Here are just a few of the trends we saw first in the Cosmos, and our ongoing efforts to respond to them:

  • Running a successful testnet is the key to a successful mainnet: We've advised the teams behind newer Proof-of-Stake protocols, like Regen Network and Solana, on how to run their own successful testnets (get in touch at if you'd like us to do the same for your team!).
  • Validator operators lack standard resources to support their work: We've built tools, like Cosmos Validator Mission Control to help make it easier for other teams to operate validators of their own.
  • Governing a Proof-of-Stake protocol democratically demands a literate, engaged community: We've remained active participants in governance across our protocols, offering feedback and sharing our thoughts on different proposals (the latest which you can see at our Governance Page).
  • Infrastructure operators are often siloed away from other teams: We've co-founded partnerships, like the Staking Defense League and the Validator Commons to hold space for indie infrastructure operators to share resources and support one another.
  • Centralization is a trend that threatens the whole point of crypto: We've written extensively over the years (here on our blog and elsewhere) to help educate around decentralization, and launched, earlier this year, Nakaflow, a tool that helps to track and compare the decentralization of stake across more than a dozen of the leading Proof-of-Stake networks.
  • Community is at the core of decentralization: We’ve sponsored events and gatherings, including this year’s Nebular Paris and ATOM Lisbon, to help support the kinds of value-aligned community that the movement for decentralization demands.

It's clear, in reflecting on these efforts, just how influential the Cosmos has been in our growth as a team. And, as we continue to grow into the new year, the Cosmos remains one of our major focuses. We're operating infrastructure on eight IBC-compatible mainnets today, and we're still getting into the dirt with new testnets.

With Game of Chains, we’re closing out 2022 as a banner year in our long history in the Cosmos. We can’t wait to see what 2023 has in store.

Eyes Peeled

Here are a few threads we’ll be keeping our eyes on over the coming weeks:

How Is the FTX Blowup Affecting Staking on Solana?

News of FTX collapsing sent the fiat price of SOL, Solana's native token, crashing. How did that crash affect staking dynamics on the network? Our friends at Staking Rewards, one of the best aggregators of data in the staking economy, shed some light on how stake has moved since November.

Also of note: in our own tracking on Nakaflow, we've seen Solana's Nakamoto Coefficient by stake, a measurement of how decentralized control of stake is across the network, hold steadily near the top of our charts throughout the FTX developments.

Crypto ❤️'s AI

With the recent launch of ChatGPT, likely the most advanced text-generating AI ever available to the public, tech discourse is, once again, buzzing about the possibilities for AI. While the current media hype-cycle around it will surely die down, AI is no doubt only going to become a bigger part of the conversation moving forward. One of the most interesting parts of that conversation is the opportunity for crypto to power the infrastructure behind future AI systems. To wit:

Akash is a protocol (which we operate on, here) that facilitates access to decentralized cloud computing. While the network's offering has to-date focused primarily on CPU-based computing power, Akash founder, and cloud-computing veteran, Greg Osuri has been previewing the protocol's forthcoming expansion into GPU-based computing power. You know what requires GPU-based computing power? AI. Some quotes from Osuri here, here, here, here, and here. That’s quite a lot of foreshadowing.

Livepeer, another protocol (which we also operate on, here) that facilitates decentralized cloud computing, has long offered GPU-based compute, primarily for streaming video applications. Livepeer CTO Eric Tang has been previewing AI on the protocol for years.

There's a running joke that capital and talent are fleeing crypto for AI, but it'll be interesting to see how these two fields can reinforce, rather than cannibalize, each other.

The Graph Turns Two 🎂

The Graph is a protocol that allows for the decentralized structuring and querying of data from blockchains, like Ethereum. This month marks the 2nd birthday for The Graph's mainnet (which we've been operating on since it first launched; find us here), and it's exciting to see the remarkable growth the network has had in that time. The Graph put together a highlight reel from different corners of the ecosystem to celebrate.

The big question on our minds heading into 2023 is: how will the protocol's transition next quarter from its centralized "hosted service" to its fully-decentralized offering go?

Liquid Staking Enters the Cosmos 🏄‍♂️

This month, we joined dozens of other validator operators in launching the mainnet for Quicksilver (find us, here), a new IBC network that enables "Liquid Staking" in the Cosmos.

What's Liquid Staking? Let's say a user wants to stake their tokens with a validator operator like Chainflow. In "Traditional Staking," the user will delegate their tokens directly to the validator. While those tokens are staked, they're effectively "locked" with the validator, and the user can't do anything with them. The tokens are, in other words, "illiquid," since they can't move while staked.

In Liquid Staking, the user doesn't delegate their tokens directly to a validator—instead, they delegate their tokens to a Liquid Staking protocol, which in turn does two things. First, it passes those tokens onto one or more validators, based the rules of the protocol (some use algorithms to select which validators receive stake, while others use hand-picked whitelists). Second, it creates what's called a "liquid staking derivative" (LSD) of the user's staked tokens. This LSD is a new type of token which acts—in theory, at least—as a claim to redeem the original tokens that the user staked. The major difference is that the LSD is "liquid," meaning it can be traded around while the original tokens remain locked in the staking protocol.

You can imagine how this might get complicated: can the LSD be used anywhere a normal token can? What happens if something breaks with the Liquid Staking protocol while your tokens are locked in it? Quite a few Liquid Staking projects have sprung up over the past couple of years, and, with the launch of Quicksilver, this emerging field has now entered The Cosmos.

Like some other Liquid Staking protocols, Quicksilver uses an algorithmic system to decide where staked tokens go. One of the considerations in that algorithm is the decentralization of stake, meaning that Quicksilver should—in theory, at least—distribute stake more evenly across all active validators over time, helping to improve decentralization in The Cosmos. While we're supporting the project out of optimism for that impact, we'll keep you posted on how things play out.

Chainflow Operations Note

'Tis the season: Chainflow will be taking a programming break from December 24th, 2022 through January 1st, 2023. We'll still be monitoring our infrastructure operations during that time and will be prepared to respond to any emergency issues, but we'll be a bit quieter on our other channels while the team takes a break after a busy year.

As always, we're sincerely grateful for your support, and wish you and yours all the best. See you in the new year.