I have successfully contributed to the @penumbrazone summoning ceremony phase 2 at slot 2698!
Hash: 6DA9D1C6B514B4C3B744CE8C17550DC5C863B97E437AF2CD2D0B5544E5642EC1
I just contributed to the @union_build Trusted Setup Ceremony, to secure its ZK circuit for trustless, decentralized interoperability.
I attest to my contribution. My public key hash is:
https://t.co/xSxFg6urj6
#JoinTheUnion
For #DevWednesday Stargaze will be periodically recognizing exemplary participation from validators on-chain and in the community.
This week we are shouting out @alkadeta validator for their 100% governance participation rate 🗳️ ✅
I have successfully contributed to the @penumbrazone summoning ceremony at slot 5768!
Hash: 5895E6699AC6F0FE12950BB493A1774B3E61860AC62DA0B3DA7D60E3508D0107
I alkadeta agree to the rules outlined in the Operating Rules of the Simple DVT Module for @LidoFinance x @ObolNetwork Node Operators in the Jubilant Jackrabbit cluster as seen by my verified signature: https://t.co/Pzdf1aNTPx
Why this tweet is absolute ****shit and shows an insufficient understanding of the Cosmos techstack its benefits and the actual acknowledged drawdowns 👇
First, lets quickly remind ourselves why @berachain (and many other teams) chose the cosmos stack in the first place:
1. ABCI - "lifecycle for blockbuilding" - In cosmos a block is not merely execution of base-layer EVM calls. A significant amount of control exists over the potential priority and moment of execution of different functions wether they are SDK native or VM related. Things like Begin and End block are powerful automation features that reduce total computational overhead and increase security and decentralisation for building systems like intent-swaps, liquidation engines, oracles and more.
These features are taken to the next level by teams like @SkipProtocol , @dYdX , @Neutron_org , @SeiNetwork and now also berachain.
2. Separation of consensus, application layer and VM/execution framework. As Berachain wanted to implement their own consensus layer reliant on their core VE-governance and liquidity-staking methods it was important that it could build this without ruining the potential implementation of Go-Ethereum on top of that. Polaris EVM as modular building component does exactly that, it runs as a plugin communicating with ABCI to the consensus layer completely separate to the non-VM CosmosSDK application layer frameworks like /Gov and in their case also /dex /oracle and so forth. This mean they can build out their own consensus framework, build out their own application layer adjustments (more scalable than forcing it in the VM) and still have a completely EVM compatible execution layer (for permissionless external devs) on top that is composable with the underlying capabilities.
3. And this composability is the third factor, not just vertically through the stack but also horizontally to other chains. The VM can interact with the application layer, which in turn can interact with consensus. Through this application layer they can also connect to any other IBC enabled blockchain and thereby have composable cross-chain contracts. This is also a major reason for @dymension to use the Cosmos stack as no version of a validium/l2 has this composability with the base-layer or other chains due to finality constraints. Using this composability Berachain, in theory, also becomes infinitely scaleable. This as separate base layers or sovereign rollups can exist with polaris to scale the total ecosystem throughput with the same validator set/trust assumptions. Interacting with $BGT and BeraDex from a smart-contract is cool in and of itself, also being able to do this with any IBC compatible blockchain is the true level-up though.
So the separation and programability of the different layers is what made the Cosmos Stack the right choice - adding to that is the existence of validators which creates a more decentralised environment (over a sequencer) where the VE-Governance of proof of liquidity can really shine.
So could BeraChain have been a rollup, i guess so - but it would not have had the features it needed to build out the incredibly developer friendly and flexible product environment they have right now.
Most of all, it would not even be more scaleable....
-----
Now on to the mentioned "scaling issues" that the berachain testnet has run into.
The testnet so far has run close to spotless:
- There is no lasting delay in block times
- There is no significant downtime of multiple validators
- There are no CometBFT gossiping issues yet like that plagued Sei and other cosmos networks in the past
- There is no breaking happenings that caused halts or other consensus related bugs.
The problems that have arrived though are:
- Validators with subpar machines falling behind (Specs are still incredibly reasonable)
- BeraChain official Artio testnet RPC cluster is getting hammered into oblivion by requests and is having downtime/scaling issues
- MemPool is getting clogged stealing resources from RPC nodes and validators alike
So what does that actually display, where does the problem lie and how have other networks fixed this.
The potential optimisations lay with the p2p layer and the design of the cometBFT mempool and free prioritisation mechanism, two aspects which have been defined by most Cosmos builders (@zmanian@jackzampolin@hxrts etc.) as the most important to fix for the whole stack to be truly robust.
Currently on the berachain testnet users don't really pay for gas as tokens are free. This causes a scenario where spamming the chain or genuine high usage (for airdrop farming, excited community testers etc) has no real costs. People keep committing transactions to the mempool, seemingly with no timeout. As the block proposers include TXs the mempool keeps filling up, a large part of these might even be incompatible transactions, which takes up bandwith and computing resources from the validators. As the mempool becomes fuller and fuller the Query nodes lose performance capacity to perform RPC queries reducing the total capacity of the cluster. Simultaneously validators with lower bandwith plans or insufficient memory speed will be unable to verify the block within the alotted time and start missing.
So what is happening is not necessarily too much transactions in the blocks, moreso too many txs waiting to be committed in a block. Validators can align on better peering settings, reduces or increased mempool size and frontends can use timeouts to almost immediately remedy that specific problem.
RPC clusters can simply add more query nodes and over time improve the software (like @terra_money did with MantleMint) to increase performance where needed.
Looking at the past 1000s of blocks we see ~1-2% of the total 30m gas block capacity is used on average. That doesn't look like a throughput issue at this moment in time.
Live L2s like the OP stack are not doing significantly more traffic than chains like @osmosiszone or @berachain, so let's not act like them porting is somehow the right move. L2s have their place, especially in reducing overhead computation cost, but for Bera it doesnt seem like the right move - especially not for scaleability reasons.
---
In the long run though improvements can and will be made.
@osmosiszone and @SkipProtocol are working together on making the blockSDK eip-1559 blockbuilding/fee prioritisation as robust as possible. This should alleviate concerns with mempool clogging as there is a real cost to making a TX and also helps support "liveness" by building specific gas lanes for functions like IBC and staking.
At the same time we have:
- builders like @SeiNetwork and @dYdX significantly improving the network p2p level (faster syncing, optimistic tx sharing, in-memory orderbooks etc.).
- Developers and CosmosOGs (@informalinc@iqlusioninc@notionaldao ) looking for change in the mempool design and reducing dDosability by clarifying potential attack vectors.
- Further improving and expanding IBC by making it fee-aware and bringing new standards to improve the dev-experience and reach of the technology ( @interchain_io@strangelovelabs@union_build@Polymer_Labs )
The Cosmos TechStack is not perfect, probably it won't ever be. But the highly flexible nature is not just a benefit for the developer experience, it also brings additional throughput. Berachain much like dYdX and the recent @CelestiaOrg launch will bring up issues Cosmos had never before imagined and test scalability. But let's not act as if some minor testnet-mempool spamming is the demise of this incredibly robust stack.
Best,
Lavender.Five Nodes 🐝💜
We're honored to be selected for the latest Evmos Foundation Delegation Program!
Congrats to all the teams who have been selected, and a big shout-out to the @EvmosOrg team for recognizing our efforts! ☄️
The Foundation Delegation Program Cycle 1 📢
After thoroughly reviewing the applications for Cycle 1 we are thrilled to announce the list of validators who will be receiving delegations from the Evmos Foundation for the next 24 weeks! 🎉
https://t.co/mWpE8LDx6v
(1/7) A significant issue since I started working on #Ethermint back in 2020 was the maintenance overhead of having to support both @cosmos and Ethereum types (blocks, transactions, gas logic, etc).
A thread 🧵
Validators & Node Runners get paid to run Evmos RPC nodes on @lavanetxyz
3 Days Left to Register
@alkadeta Thank you for compiling this information 🙏
Learn how 👇
@ShrimpClubNFT@Gray0024@stride_zone@cosmoshub This is not a Stride Proposal. These are all Stargaze validators, and this image is taken from a Stargaze proposal. Most of the validators in this image, including us don't even validate on Stride.
@EvmosOrg 2.0 is the realization of the original vision of a fully interoperable #EVM, and will remind everyone why we’re the team behind the original #EVM on @cosmos
👉 New Evmos Tokenomics will improve elasticity in the token emissions and distribution to adapt to market conditions and provide incentives for growth
👉 Evmos EVM Extensions will bring massive innovation to @cosmos by enabling #StakeFi and IBC composability for smart contracts
👉 Outposts (@stride_zone and @osmosiszone) are going to expand the limits of liquidity by enabling dApps to connect directly with IBC chains
👉 Dynamic IBC (dIBC) will enable smart contracts to create their own IBC apps, enabling the next generation of dApps to be built on Evmos
👉 A new Software License that will protect and defend our IP rights and create a direct value alignment between Evmos SDK chains and our Evmos users
👉 Finally, our long-awaited rebrand to mark this new phase in our project under a leadership team that is 💯% committed to its success
Onwards and upwards 🫡