Implemented Based Contestable Rollup (BCR) design with multi-proofs and SGX is the main change from previous Alphas: “With Katla, we enable a multi-proof configuration with four tiers: Optimistic proofs, SGX proofs, SGX+zkEVM proofs, and guardian proofs.” Read more about BCR here:Twitter Thread and Blog writeup by Daniel Wang
Katla’s code supports EIP-4844 (proto-danksharding) but this feature is currently turned off. (This will be an upgrade to Alpha-6, to-be enabled when Holesky enables its support for EIP-4844)
All of the Taiko alphas have been named after Icelandic volcanos. Katla is one of these Iceland volcanos.
For EIP-4844, it is looking like it will be deployed across all 3 testnets in a rolling fashion - on Goerli just yesterday but with some issues, Sepolia at the end of January (31st), and then on Holesky in early February (as early as the 7th).
This is part of the overall Dencun upgrade, which (all going well with the testnets) will be then deployed to Ethereum mainnet. Easy to see that this Dencun upgrade is focused on some enhancements for Ethereum in terms of optimising for Layer 2, with some very exciting potential for early 2024 ETH!
ie:
“This update is expected to drastically reduce transaction confirmation costs by 10 to 100 times for Layer 2 solutions, thereby lowering expenses for users within Ethereum’s Layer 2 space.”
you say such cool words that I want to re-read this many times. I love Ethereum and the more practical this blockchain becomes, the more interesting events will develop, of course, for those who are directly involved in it. for everyone else, this will also turn out to be interesting and there may be a radical increase in blockchain in society and this is a very exciting moment for the entire transaction system on the planet. This kind of information inventory capability has never existed before and it’s cool.
thanks for the answer, but when I wanted to know what this word means, the search engine volcanoes showed me, although I wasn’t sure, but now I’m sure that it means volcano. Iceland is a very beautiful country and very rich in very cheap energy and this is also a very cool moment in the history of mankind, if of course we go the right way then we will learn to tame volcanoes and this is a very good ability for civilization, because it increases the chances of survival on this planet
Next up for the upgrade is a Sepolia testnet deployment at the end of the month, and then Holesky in early February (which is when Taiko will potentially be able to test these new Ethereum upgrades).
EIP-4844 could be rolled out to mainnet as early as March, all going well. That would be a great step forwards for Ethereum and a big win for Layer2
The multi proof system means that each new block has a certain likelihood of being assigned to a proving method. But why assign 90% to the optimistic method and 10% to SGX? Why not let the market itself decide the ratios?
If it’s a matter of incentives, why not simply reward more the most secure methods? Following the same R/R Taiko bond model: the less secure the method, the more tokens it lbonds (like a collateral)
“The percentages in the below diagram represent the probability that a block will require a given tier as its minimum proof level.”
So I believe the correct way to look at this is that the Optimistic level has a 89.8% chance of being the minimum level required, SGX tier has a 10% chance of being the minimum level, SGX + zkEVM is a 0.2% chance of being the minimum (meaning that SGX and optimistic proofs are still required if this tier is the minimum).
So basically, for this testnet SGX proofs are going to be needed for 10% of transactions overall, and an Optimistic proof will be required all of the time, while 89.8% of the time Optimistic proofs are the only proof required.
Digging into this deeper, shows that with the ‘free market’ mindset Taiko is going for - proof levels could be up to users/provers to request to some degree as well. So in future you could (maybe) require SGX levels for all proofs if you personally wanted, ie this could be up to users or developers to configure as they require (altho this is starting to go outside of my knowledge on this so someone correct me if I am wrong - or it might just be too early to know for sure how this will work yet).
If it’s a matter of incentives, why not simply reward more the most secure methods?
One reward I have read about on this topic is a naturally forming one - which is simply that the most secure and proven methods will be rewarded more by nature, as they will of course be utilized more often.
We also should remember that this is an alpha testnet - so these requirements are likely set just for testing purposes and I expect will change considerably as time goes on and different setups are tested and optimized.
i add sgx may also not be available to all nodes and if you put it in first place it will affect TPS, It turns out that the optimistic level is more democratic and productive, while thanks to sgx the entire network can follow the vitalik trilemma. be secure, fast and scalable. each transaction thus receives what is required for its implementation and this satisfies the market. in a word, it’s just optimization of resources, most likely, I’m not very good at it either, but I think something like this about it all and it’s hard for me to even imagine what will happen next with blockchains
After successfully being rolled out and tested across all testnets, Dencan upgrade was just launched today on Ethereum!(here is the first blob submitted)
We now have blobs!
Theres a good writeup here covering this news:
And so what is the big deal with Dencun/EIP-4844?
“EIP-4844 introduces a new kind of transaction type to Ethereum which accepts “blobs” of data to be persisted in the beacon node for a short period of time. These changes are forwards compatible with Ethereum’s scaling roadmap, and blobs are small enough to keep disk use manageable.”
With rollups looking pretty solid as the future of Ethereum, this update will fundamentally help optimize Layer2s, and therefore greatly assists Ethereum scale overall. This means lower fees (potentially by an order of magnitude), future-proofing and setup for future upgrades, streamlined data pruning & management, faster networks (100,000-millions of TPS), and overall offers more incentives for the Ethereum ecosystem to move to rollups.
Sharding/Danksharding/Proto-Danksharding
When it comes to this update launched today, we need to understand ‘Sharding’ - a technique which has been developed to increase the efficiency/scalability of a blockchain network. It involves dividing the network into smaller units called ‘shards’, letting each shard process the transactions in parallel rather than having all nodes in a network process transactions one at a time.
‘Danksharding’ is then a simplified version of Sharding, which brings a new data layer to Ethereum (ie for rollups to utilize to store their compressed transaction data) and Proto-Danksharding (as we now have with EIP-4844 launching) is the majority of the logic and foundation for Danksharding itself.
What exactly is a ‘blob’?
Proto-Danksharding (EIP-4844) adds a new transaction type known as a ‘blob-carrying transaction’ and a new data layer for blob storage.
Blobs are 4096 field-elements of 32 bytes each, with a long term maximum of 16 blobs per block. 4096 * 32 bytes * 16 per block = 2 mb per block maximum. This new data layer exists only on the ‘Beacon nodes’ of the Ethereum Beacon Chain. The Beacon Chain is the core component of Ethereum 2.0 (launched in September 2022 when Ethereum switched to Proof of Stake with ‘The Merge’) and does not process transactions or handle smart contract interactions (this is handled by the execution layer) but instead it coordinates the consensus among validators, manages validator staking, and organizes the shard chains.
Blobs will also be ‘pruned’ after a certain time has passed (~2-4 weeks). So not only is this a more efficient allocation of extra data storage for Ethereum (as L1 block space is expensive for data storage), but this blob data is also not held onto indefinitely. The main purpose of this extra blob data is for fraud proofs, so this data only needs to be around for long enough for dispute periods to pass.
Taiko and Blobs
Yes, there are plans for Taiko to take advantage of these benefits, we could very well see Alpha-7 setup soon to test EIP-4844, or it might be enabled for Alpha-6 (we will need to wait to find out!). It will certainly be interesting to see the kinds of benefits Taiko can find here, and for us to see how cheaply and quickly Layer 2s can now operate. A big win for the Ethereum ecosystem overall.
fantastic fantastic fantastic good presentation, but as they say, it’s better to try once than to hear a hundred times, this is a super cool update for all l2 so lesgo taiko
As mentioned (and as we heard in the most recent Tea with Taiko) - it has been confirmed there will be an Alpha-7 testnet (Hekla). This testnet will likely implement EIP-4844. Check out my previous comment in this thread (above) for a breakdown of what EIP-4844 is bringing to the ETH ecosystem.
The Alpha-7 testnet is now live and open to the public, this new testnet is named ‘Hekla’! The Alpha-6 Testnet is to-be-deprecated May 10th: tweet
For continued discussions heading into the future, be sure to check out the new Alpha-7 forum post!
New Alpha forum post continued here:
Note that with the Alpha-6 testnet being deprecated early in May, you will want to bridge your ETH back to Holesky, to avoid losing it (the bridge can be found here)
Note that with the Alpha-6 testnet being deprecated early in May, you will want to bridge your ETH back to Holesky, to avoid losing it (