gakonst
0 Suivis
2 Abonnés
2 J’aime
0 Partagé(s)
Publications
Vidéos
If you're a Reth engineer and you're working with recruiters, feel free to just reach out to me directly.
We created the Reth Engineer role as one of the most important roles for the future of crypto infrastructure, and we have tons of roles in the portfolio. I'll intro you.
We created the Reth Engineer role as one of the most important roles for the future of crypto infrastructure, and we have tons of roles in the portfolio. I'll intro you.
Can you create a stablecoin out of bitcoin hashrate derivatives?
Users shouldn't pay gas fees, apps should.
NextJS example for fee sponsoring in Porto just landed (link in thread)
NextJS example for fee sponsoring in Porto just landed (link in thread)
Frontiers by @Paradigm update! Aug 6-8 SF.
Day 2 is looking 🔥🔥🔥 .
Not 1, not 2, but 3 talks on high performance!
- "Hyperoptimizing Reth" by @ashekhirin & @Rjected
- “Scaling Ethereum's State Trie” by @brianisbland
- "Scaling Ethereum L1" by @adietrichs.
Apply below!
Day 2 is looking 🔥🔥🔥 .
Not 1, not 2, but 3 talks on high performance!
- "Hyperoptimizing Reth" by @ashekhirin & @Rjected
- “Scaling Ethereum's State Trie” by @brianisbland
- "Scaling Ethereum L1" by @adietrichs.
Apply below!
What's the fastest folding / IVC SNARK atm?
Does it have an optimized implementation?
Does it have an optimized implementation?
Still looking for a consensus expert on stake-based finality gadgets for proof of work blockchains, bonus points if you have studied proof of useful work / MEV market structure etc
would be awesome if my crypto social graph kept it technical when things are happening in the world or stuck with democratic and peace making takes, just better given crypto is supposed to be tech for global freedom and peace
Things that are ultimately not gonna help Ethereum win for Glamsterdam: EOF, EVM64, SSZ/pureth, Available Attestations.
Need strong offense, not weak offense, and definitely not long term defense atm. Short term defense (eg repricings) are good.
Need strong offense, not weak offense, and definitely not long term defense atm. Short term defense (eg repricings) are good.
RE: what should happen to Ethereum's EL my current view is:
- Fusaka scores low hanging fruit wins and sets caps for raising the gas limit.
- Glamsterdam continues the above gas limit bumping while repricing to bound worst case outcomes.
My crazy take might be that after Glamsterdam, beyond further repricing and gas limit increases we need to switch to optimizing the EL for ZK usage, whatever that may mean.
- Fusaka scores low hanging fruit wins and sets caps for raising the gas limit.
- Glamsterdam continues the above gas limit bumping while repricing to bound worst case outcomes.
My crazy take might be that after Glamsterdam, beyond further repricing and gas limit increases we need to switch to optimizing the EL for ZK usage, whatever that may mean.
Fusaka will happen this year and enable the L2s to further scale.
Blobs will stay the same on launch, and with pre scheduled blob only forks they will go up to 48 hopefully around Q1'26.
We will keep measuring and see how high we can go, tools are there, just reps now.
Blobs will stay the same on launch, and with pre scheduled blob only forks they will go up to 48 hopefully around Q1'26.
We will keep measuring and see how high we can go, tools are there, just reps now.
we are getting better at using AI to improve our productivity and empower the open source community.
in this PR we share the prompts where Claude successfully built a low level EVM call checker lint for Forge Lint (runs by default on forge build)
https://github.com/foundry-rs/foundry/pull/10810
in this PR we share the prompts where Claude successfully built a low level EVM call checker lint for Forge Lint (runs by default on forge build)
https://github.com/foundry-rs/foundry/pull/10810
What I think is important for Ethereum to win while maintaining a global node set:
1. decouple validation from block building throughout with zk. this needs making zk proving blocks in real time easier: ePBS or Delayed Execution, I don't care which. trie migrations help but imo not needed, whereas ensuring you don't prove at the end of the slot but at the start really matters.
2. decouple censorship resistance from weakest node. If you do the above, introducing low-ETH staked nodes w FOCIL who are responsible just for CR but not for the hot path block building and validation liberates us from raspberry pis for scaling but let's us continue using raspberry pis for CR
1. decouple validation from block building throughout with zk. this needs making zk proving blocks in real time easier: ePBS or Delayed Execution, I don't care which. trie migrations help but imo not needed, whereas ensuring you don't prove at the end of the slot but at the start really matters.
2. decouple censorship resistance from weakest node. If you do the above, introducing low-ETH staked nodes w FOCIL who are responsible just for CR but not for the hot path block building and validation liberates us from raspberry pis for scaling but let's us continue using raspberry pis for CR
What I think is important for Ethereum to win while maintaining a global node set:
1. decouple validation from block building throughout with zk. this needs making zk proving blocks in real time easier: ePBS or Delayed Execution, I don't care which. trie migrations help but imo not needed, whereas ensuring you don't prove at the end of the slot but at the start really matters.
2. decouple censorship resistance from weakest node. If you do the above, introducing low-ETH staked nodes who are responsible just for CR but not for the hot path block building and validation liberates us from raspberry pis for scaling but let's us continue using raspberry pis for CR
1. decouple validation from block building throughout with zk. this needs making zk proving blocks in real time easier: ePBS or Delayed Execution, I don't care which. trie migrations help but imo not needed, whereas ensuring you don't prove at the end of the slot but at the start really matters.
2. decouple censorship resistance from weakest node. If you do the above, introducing low-ETH staked nodes who are responsible just for CR but not for the hot path block building and validation liberates us from raspberry pis for scaling but let's us continue using raspberry pis for CR
I'm still baffled that the Ethereum Core Dev community does not prioritize fixing the 2 most cited problem of EVM developers per the Solidity Lang survey despite our repeated efforts:
1. Stack too Deep: yes this is a Solidity skill issue a little bit but just add a SWAP/DUP17-32 opcode range and call it a day. You will burn some opcodes. It's fine, they are meant to be used. You're gonna have another PUSH0-style mismatch, this is also fine, it's not perfect but it's fine.
2. Lift the 24KB limit. I don't really care what you do, make it 32KB, 48KB, 128KB, 256KB, 512KB, do it all at once, incrementally, price it or not but do something! Now, not next year!
If you are scaling the L1, ensuring people can write contracts without stupid errors is P0.
If the system cannot handle an extra 8KB per bytecode which is a param that was set 10yrs ago literally then there's no chance you will be able to actually scale the L1.
Fix stack too deep and bytecode size limit! For the devs!
1. Stack too Deep: yes this is a Solidity skill issue a little bit but just add a SWAP/DUP17-32 opcode range and call it a day. You will burn some opcodes. It's fine, they are meant to be used. You're gonna have another PUSH0-style mismatch, this is also fine, it's not perfect but it's fine.
2. Lift the 24KB limit. I don't really care what you do, make it 32KB, 48KB, 128KB, 256KB, 512KB, do it all at once, incrementally, price it or not but do something! Now, not next year!
If you are scaling the L1, ensuring people can write contracts without stupid errors is P0.
If the system cannot handle an extra 8KB per bytecode which is a param that was set 10yrs ago literally then there's no chance you will be able to actually scale the L1.
Fix stack too deep and bytecode size limit! For the devs!
I'm still baffled that the Ethereum Core Dev community does not prioritize fixing the 2 most cited problem of EVM developers per the Solidity Lang survey:
1. Stack too Deep: yes this is a Solidity skill issue a little bit but just add a SWAP/DUP17-32 opcode range and call it a day. You will burn some opcodes. It's fine, they are meant to be used. You're gonna have another PUSH0-style mismatch, this is also fine, it's not perfect but it's fine.
2. Lift the 24KB limit. I don't really care what you do, make it 32KB, 48KB, 128KB, 256KB, 512KB, do it all at once, incrementally, price it or not but do something! Now, not next year!
If you are scaling the L1, ensuring people can write contracts without stupid errors is P0.
If the system cannot handle an extra 8KB per bytecode which is a param that was set 10yrs ago literally then there's no chance you will be able to actually scale the L1.
Fix stack too deep and bytecode size limit! For the devs!
1. Stack too Deep: yes this is a Solidity skill issue a little bit but just add a SWAP/DUP17-32 opcode range and call it a day. You will burn some opcodes. It's fine, they are meant to be used. You're gonna have another PUSH0-style mismatch, this is also fine, it's not perfect but it's fine.
2. Lift the 24KB limit. I don't really care what you do, make it 32KB, 48KB, 128KB, 256KB, 512KB, do it all at once, incrementally, price it or not but do something! Now, not next year!
If you are scaling the L1, ensuring people can write contracts without stupid errors is P0.
If the system cannot handle an extra 8KB per bytecode which is a param that was set 10yrs ago literally then there's no chance you will be able to actually scale the L1.
Fix stack too deep and bytecode size limit! For the devs!
I'm still baffled that the Ethereum Core Dev community does not prioritize fixing the most cited problem of EVM developers per the Solidity Lang survey
1. Stack too Deep: yes this is a Solidity skill issue a little bit but just add a SWAP/DUP17-32 opcode range and call it a day. You will burn some opcodes. It's fine, they are meant to be used. You're gonna have another PUSH0-style mismatch, this is also fine, it's not perfect but it's fine.
2. Lift the 24KB limit. I don't really care what you do, make it 32KB, 48KB, 128KB, 256KB, 512KB, do it all at once, incrementally, price it or not but do something! Now, not next year!
If you are scaling the L1, ensuring people can write contracts without stupid errors is P0.
If the system cannot handle an extra 8KB per bytecode which is a param that was set 10yrs ago literally then there's no chance you will be able to actually scale the L1.
Fix stack too deep and bytecode size limit! For the devs!
1. Stack too Deep: yes this is a Solidity skill issue a little bit but just add a SWAP/DUP17-32 opcode range and call it a day. You will burn some opcodes. It's fine, they are meant to be used. You're gonna have another PUSH0-style mismatch, this is also fine, it's not perfect but it's fine.
2. Lift the 24KB limit. I don't really care what you do, make it 32KB, 48KB, 128KB, 256KB, 512KB, do it all at once, incrementally, price it or not but do something! Now, not next year!
If you are scaling the L1, ensuring people can write contracts without stupid errors is P0.
If the system cannot handle an extra 8KB per bytecode which is a param that was set 10yrs ago literally then there's no chance you will be able to actually scale the L1.
Fix stack too deep and bytecode size limit! For the devs!
Every piece of crypto infra will be touching Reth one way or another.
The most important thing for it's success is by far the OSS community. A trained group of >500 geo distributed people that have contributed to Reth and are bought into the vision and its long term success.
On a technical level, the Reth project has 3 pillars:
- Security: We get that by supporting Ethereum L1 in production staking environments. Skin in the game.
- Performance: We get that by pushing the frontier on L2s and MEV Block building. Terragas.
- Extensibility: We provide excellent APIs for modding your node without forking. Reth SDK.
So much more to do, but really proud of where we are today.
The most important thing for it's success is by far the OSS community. A trained group of >500 geo distributed people that have contributed to Reth and are bought into the vision and its long term success.
On a technical level, the Reth project has 3 pillars:
- Security: We get that by supporting Ethereum L1 in production staking environments. Skin in the game.
- Performance: We get that by pushing the frontier on L2s and MEV Block building. Terragas.
- Extensibility: We provide excellent APIs for modding your node without forking. Reth SDK.
So much more to do, but really proud of where we are today.
If Ethereum goes 100% ZK it makes most sense to treat the provers like miners and pay them issuance.
That was my original thesis for ZK Hardware and Ethereum is the best place for this to happen as it has deep liquidity and valuable issuance vs any new ZK L1.
That was my original thesis for ZK Hardware and Ethereum is the best place for this to happen as it has deep liquidity and valuable issuance vs any new ZK L1.
Connectez-vous pour découvrir d’autres contenus