Thread by @buchmanster

@cosmos is not a single thing or single network. It's a philosophy of blockchain design that has materialized in major advances, including some of the most useful software artifacts available today to blockchain developers. Let's take a look at what Cosmos has created:

https://twitter.com/jeffscottward/status/1247741371051585536

1) Tendermint. Undisputed world champion production consensus engine. No other system comes close (yet!). Lots of production non-BFT systems out there (@etcdio , @hashicorp #consul, etc.) but no BFT system has been battle tested as much as Tendermint.

If you believe there will be more than a couple blockchains out there (see later), then production grade BFT matters. Besides, Tendermint BFT has been massively inspirational and influential in practically every other emerging blockchain consensus, including ETH 2.0 Casper.

2) ABCI. With ABCI (Application BlockChain Interface), you can build applications on Tendermint in *any* programming language. You know, like one's you've been writing in for decade/s with real dev environments and tooling and testing stories and massive communities.

ABCI makes blockchain development enormously accessible. But it also facilitates running existing blockchain apps (eg. the EVM itself) on a production grade BFT engine - eg. @ethermint (EVM on Tendermint). So you get the best of both worlds!

3) Cosmos-SDK. The Ruby-on-Rails for blockchain dev. Let's you build new production grade blockchain apps on Tendermint rapidly. Experience reports suggest that building with the Cosmos-SDK is 10x more productive than building on the EVM.

This might be almost entirely due to the developer tooling gap between Go and Solidity. But the Cosmos-SDK promises more than that, it promises interoperability through ...

4) IBC. If there's going to be many blockchains (there will), we need protocols to enable them to communicate. That's IBC. But IBC is a general purpose protocol - it can in theory even be used to define the ETH 2.0 sharding protocol (!).

The current focus is on using it to connect the growing ecosystem of @cosmos blockchains - @cosmoshub, @agoric, @regen_network, @AltheaNetwork, @emoney_com, @irisnetwork, @Binance_DEX, @akashnet_, @cyber_devs, @corestar_io, @loomnetwork and more!

Various people are also working to various extents on integrating with other PoS protocols via IBC, like @polkadotnetwork, and with existing blockchain networks like @bitcoincoreorg , @ethereum , and @ZcashFoundation

5) Validators. If you're a serious PoS validator today, you most likely cut your teeth in a Cosmos testnet. If you're launching a PoS network today, you're probably running an adversarial testnet. Cosmos pioneered all of that. ETH2.0 validators will be better thx to Cosmos

If nothing else, the point of @cosmos was to push the *entire* industry forward with practical incremental deliverables. @zmanian does a great job describing our GoToMarket strategy over the last few years in his MIT BitcoinExpo talk:

Ok, that's great you might say, and thanks Cosmos. But why should we accept that there will be more than one blockchain at all? If ETH2.0 works, and the world-computer vision is attainable, why would we need that?

Well, the World Computer vision is a misnomer ...

It's about as promising as a "World Mainframe" vision 30 years ago. There's 1000 reasons why we had to move away from single large mainframes (even "scalable" ones) to more decentralized networks of machines interoperating via communication protocols (TCP/IP).

And all those reasons will apply to ETH 2.0: latency, scalability, governance, customization, etc. You're simply not going to design a single system that satisfies the complete computing needs of the entire world. It's absurd to think you can.

And just like mainframes needed to break up into distributed heterogenous machines interoperating via comms protocols like TCP, so too will big blockchains need to be broken up into distributed heterogenous blockchains (abstract machines) interoping via comms protocols like IBC

Not to say ETH 2.0 isn't a worthy project. It's an incredible project that is undoubtedly pushing forward the state of the art. But the @cosmos vision is about sovereignty and sustainability. It's about #localism and freedom of choice. It's about composable heterogeneity.

It's about being there where ETH cannot. Not because ETH failed, but because the design space is too big. ETH is a top-down project. Cosmos is bottom up. We will meet in the glorious middle, and all be better off because of it.

@cosmos is the technology stack that puts the application developer and community in charge, but preserves the ability to connect with others. Today it might be difficult to find a secure validator community for an app specific blockchain. But it's getting increasingly easier.

More folks are setting up PoS companies. New protocols like cross-chain staking over IBC and @lazyledger_io make it possible to leverage existing validator sets for new chains. But companies and jurisdictions are increasingly taking their tech stacks into their own hands.

@cosmos lays the foundation for local currencies and for every city to run their government on a blockchain. They won't all just run on ETH2.0, because the env is too far out of their control. ETH2.0 is for global applications. But the future will be increasingly localized ...

... and the @cosmos philosophy of blockchains is #localism. /end :)

If you're looking for more on the philosophy of @cosmos, check out my "Stakeholders and Statemachines" talk from last year at

For more on the past/present/future of cosmos, see my recent blog post: ebuchman.github.io/posts/2020-03-…