Great guest for our OCLUB AMA episode 4, ‘A Deep Dive into StarkWare’, we had none other than Liron, Business Development Manager at StarkWare. StarkWare’s exciting updates, Cairo, StarkNet’s roadmap so and so forth were covered in this session.
Click to play our AMA recording on Spotify.
- Liron’s journey into crypto world and introduction of StarkWare’s team. [01:25]
Liron: On a personal level, I’ve been with StarkWare for roughly 2 years, always been a blockchain enthusiast probably since 2017. And Louis you’ve probably seen on Twitter discovered me on his telegram chat asking questions about automated market makers. And that really got me the interview at StarkWare. For the team in StarkWare, the founding team invented these STARKS, which I will get to it later there. They have a very technical, photographic academic background. We have 100 or so team members, mostly part of the engineering building up this technology to support STARKS on the blockchain. It’s in Tel-Aviv.
2. Highlights in StarkWare Session 2023 and updates on StarkNet on Mobile. [03:08]
Liron: We had over 1,000 attendees in person. The majority were engineers, developers, very positive energy, very exciting. There were a few different tracks. There was a research track, the workshop track, just generic talks. If I would summarise the conference in one word, would be open source, I guess. A lot of announcements at the conference which is also good start to the open source. It’s just like a very positive community vibe towards StarkNet, towards scaling Ethereum with the zk-rollups. Specifically with StarkNet on mobile, there’s a lot of really good efforts going on infrastructure side of things. One of the great things is about building permissionless, decentralised blockchain infrastructure. So building a lot of the blockchain core infrastructure, like the sequences and the cryptography. But the infrastructure surrounding that license, the mobile support comes mostly from wallet or providers.
3. Definition of the Three ‘S’: STARKS, StarkEX, StarkNet. [05:54]
Liron: The STARKS, it’s an abbreviation for a very dynamic terms, scalable, transparent, argument of knowledge, which is a very fancy way of describing the proof system. What’s unique about this STARKS cryptographic proof system, I wanted to make this understandable for the people listening to this call, even if you’ve never really at exposure to STARKS before is that this imbalance here between how easy it is to verify? The proof that the STARK generated. So let me start with the analogy. When you have your bank balance and you have 10,000 transactions in your bank balance. And as a result of this, it has a 10,000 transactions, including some incoming transactions on how to introduce that means your final balance is U.S.$1,000. Now, if you want to go to somebody else to verify your balance that person has to process all of those 10,000 transactions and verify themselves and your final balance of the US dollars. We call that linear verification, meaning the verifier has to run the same number of transactions that you processed. With STARK proofs, what happens here is through the magic of this cryptography that our founders are pioneers. The verification only requires doing 100 checks rather than 10,000 checks. That’s a fundamental innovation behind STARKS, and a fundamental innovation that the company is driving. This innovation of STARKS to the verification is exponentially smaller than the normal transactions. The primary use cases in today’s world is in the blockchain space. As you probably all know, when Ehereum is congested, to either scale or transaction cases become so exorbitant and expensive. That is not feasible for your transaction on chain. Instead of submitting individual transactions to the blockchain, you can submit groups and the tests to the validity of multiple transactions. Therefore, what the block chain does is verified proofs, because its verification is so much more efficient than running the individual one, you get low transaction costs while scaling the blockchains.
That’s the application that STARKS come for. Where we will apply, that is in two different products. The first one is StarkEX. The second one is StarkNet. StarkEx is what we call an APP chain as a service product, where you engage in a business relationship with StarkWare and we build a dedicated APP chain for your own purposes where you run your transactions on that APP. There are some examples, like Immutable X running an NFT marketplace. This so rare, doing an NFT marketplace or a sports, mainly baseball and Premier League, a bunch of sports. Dydx, a potential trading platform. It does over $1 billion a day. StarkEx can really be done like over 300 million transactions and over billions of dollars of volume. It’s a very good APP chain products that works very well. So basically what happens is the transactions that take place in the StarkEx environment get sent to our STARK prover. Regenerate the proof and we submit that proof to Ethereum. And NFTs on Immutable today, for example, is roughly 1,000 times cheaper than it is to do it natively on Ethereum. StarkNet is a permissionless open blockchain which will be decentralized in the near term future. With StarkNet, it works like any other public blockchain where you can permissionlessly write a small contract and deploy a smart contract on chain, then we get it over there behind the scenes. We take care of generating a proof and submit that proof to Ethereum. There’s lots of addresses but just to summarise STARKS, all these are cryptographic proofs that we generate. StarkEX is the APP chain product that StarkWare runs, StarkNet is the public permissionless blockchain that is where STARKS buildings and infrastructure for.
4. Share Cairo from a beginner-friendly way. [17:15]
Liron: We’re trying to design as described before, and we chose to write this language called Cairo. Let me talk historically. Before Cairo, when we wanted to prove something, we would have our photographers and engineers sit in a room with whiteboards and write down algebraic statements and turn the whole day — just generating these algebraic statements to prove. Very tedious and very prone to human error. So internally, they built this tool called Cairo to scale that out, and we built it internally. It was low level like writing it in assembly or machine-level code. Once that was mature, they realized it was worth externalizing this code and making it the code language for StarkNet. We’re currently going through an upgrade to go. And this evolution to Cairo basically makes it syntactically very similar to Python. You never even start like algebra polynomials when writing your code. This new version of Cairo supports syntax. Interested datatype is one thing that you’re probably familiar with just from either language. That doesn’t make the developer experience very similar to writing in Python, so it makes it accessible to developers. And that’s a basic overview of Cairo.
5. Toss about the evolution of Cairo 1.0 and Starknet 0.11.0, what’s next? [19:52]
Liron: So they can get into technical stuff trying to keep the high level. Cairo1.0 has a bit of work to be done to get the full feature parity. Cairo 0.9, that ETA is like, let’s say, end of the month or a few weeks. The next thing is to make Cairo1.0. The programming language can compile into the StarkNet operating system. So StarkNet 0.11.0, which is the next release of StarkNet that, will be released to support Cairo1.0, which is a very big deal, which is the new smart contract. We have to have the code for another language. This will also look technical, but every release of StarkNet will be linked to a specific compiler. So like starting to promote because there will only work with Cairo1.0 of altering three releases. I think to become more stable, that should be some backward compatibility that. For you, an end developer, the main thing to do is look at the data types and the tutorials exercise because that stuff should not change so much as we add features.
6. Validity rollups are not limited in throughput in the same manner as L1s. This gives rise to potentially much higher TPS on L2 validity rollups. StarkNet as a permissionless decentralised Validity-Rollup, could you please share StarkNet performance roadmap? [22:14]
Liron: Our blog was posted in late November describing the performance roadmap, and a few things there have been implemented. Some things are still ongoing. For example, they are about the ability to parallelize transactions. It’s actually live already. That’s a big breakthrough. So explain what this means. Let’s say I’m sending you money while simultaneously Buri is sending money to Ellie. These two transactions are completely independent. My balance of ETH doesn’t impact ordinariates’ ability to transact. What parallelization transactions enable is that as long as transactions do not affect the same state. They can be run in parallel sequence which provide a lot of scalability benefits.
The other performance improvements that are planned or ongoing are basically moving both the sequencer and the virtual machine to Rust, which is, again, quite technical but quite important. The sequence is being Rust. These provide very interesting benefits to the performance of StarkNet. Specifically, there’s also an additive benefit in not only the sequence of being Rust but also the virtual machine being Rust. These are both in the Rust language.
There are actually other benefits. So that’s basically the ongoing performance now, specifically with the Cairo1.0 roadmap. It basically works like this, your smart contract was written in the previous version of Cairo and is now not supported by the new STARK. So you have to go through an upgrade to this, which means that if you had already deployed code from the previous version of Cairo on chain, you have to just upgrade your contract. And obviously, we don’t want people to deploy previously to lose access to their funds or the controls of the smart contracts. There’s a period, I guess a transition period, where you’re able to upgrade your contracts. It also lasts a few months, and then after that period is over, there’ll be what’s called a regenesis blockchain where there’s a snapshot taken over contracts. And then, with that will, you’ll only be able to replay from Cairo1.0, and people who do not upgrade from Cairo will probably miss out. There’s quite a lot going on. Once that is stable, then this decentralization class, etc.
7. What kind of mental models that EVM developers should keep in mind in order to build on StarkNet? [31:24]
Liron: Stop thinking within the EVM constraints. There are the compute power and reduction cost for compute that the zk-rollups enable, specifically StarkNet, allowing you to think of stuff that will it wasn’t previously possible on EVM. Beyond computing, there are also a lot of technological unlocks. One example is account abstraction, which is a buzzword that people mention it all the time. And I’m just there is a concern of mine at some EVM, developers don’t, you know, they need to get that because they’ve been in the EVM for so long. That was possible with account abstraction, which isn’t currently natively supported on Ethereum. Hopefully, with a few proposals, they will become supported. So those would be mental models, like thinking outside of EVM. Think about stuff that’s the heavy computation of what you wanted on the blockchain and figure out what’s possible now, like account abstraction in terms of actual, like, verticals. I think something like gaming is a really good example for people interested in gaming. Until now, a lot of gaming blockchains have really just evolved. Maintaining assets like NFT is of the end game assets and then transferring and trading another blockchain but the actual end game logic needed to be secured on the blockchain. Now with cheap computing that the developers enable, you can actually do the game logic on chain as well.
8. How can account abstraction actually improve security and user experience for mainstream crypto adoption？[33:32]
Liron: I just want to make this clear for everybody on the call. It’s even less technical. Ethereum basically divides contracts into contract addresses and externally owning addresses (EOA), so when you have your address in MetaMask, that’s not a smart contract. That is just an address, and you can receive funds. It can abstract and basically makes your wallet when you receive funds itself can be a smart contract, and therefore, your wallet can have functionalities that you never thought were possible. Instead of calling it account abstraction, I call it a smart wallet so that the average person can understand what’s being unlocked. One example is recurring payments. So like, I have a Spotify account and may pay $12 monthly, every month they charged my credit card, and my credit card authorisation paid the money. In the blockchain world, there’s no real paradigm right now for recurring pool transactions. You have to sort of lock up the funds for the next 12 months in a smart contract. And then they’re released from that smart contract every month, but then you haven’t really got the benefits of monthly payments because the annual opinion is already locked up in day one. So now account abstraction like recurring payments where you say in your hot wallet, your smart wallet, you authorise the people to pull funds from you, the right people to pull funds every month. You can basically top up your wallet the minute before the payments are due, and then the payment can be put in.
Visa is one example of a theme that’s uncovered how to track StarkNet specifically for this use case. And they’re building on StarkNet not just because of the security or the validity or the account abstraction. One big risk these days is how to secure a private key with the account abstraction. Your funds can be released if you login successfully using Face ID on your iPhone in order to release your funds and therefore your biometrics can secure your private key, so you’re not worrying about storing a 12-word or 256 bits on a piece of paper. Those are some use cases. I think it was a really good user experience, upgrades to all of us.
Starkware develops STARK-based solutions for the blockchain industry. StarkWare products facilitate secure, trustless, and scalable blockchain applications.
About Orbiter Finance
Orbiter Finance is a decentralised cross-rollup bridge for transferring the Ethereum native assets, which is the infrastructure of Layer 2, we offer low cost and almost instant transfers.