Eric Ng

Most Cairo devs still write while loops. But iterators in Cairo are not just cleaner, they’re cheaper. In benchmarks, iterator-based code is 20–30% more gas-efficient than equivalent loops. I wrote a full deep dive with examples, patterns, and benchmarks👇

As someone who has tried v0, lovable, and then this, hear me out. If you are building any UI, this is what you have been waiting for, and it is free (for now)

@maraoz Good read, really appreciated

Request for Human Intervention maraoz.com/beto/

@predmarks para empezar, donde cuando dices va a pasar esto significa que de verdad lo crees

Las encuestas no alcanzan. Podemos tener algo mejor.

@predmarks sorprende que todavia no haya uno serio

Argentina necesita prediction markets Ya nadie cree en las encuestas, y entre el ruido de las redes y las fake news, es imposible informarse.

These will also be available soon in the Cairo Wizard 🧙‍♂️ wizard.openzeppelin.com/cairo

New alpha for OpenZeppelin Contracts for Cairo is out! 🚀 Some highlights I’m pretty excited about: - ERC-6909 standard + extensions - ERC20 & ERC721 Wrapper components - ERC721Consecutive + BitMap utilities Full details here 👇 github.com/OpenZeppelin/c…

@0xprivacypools @Starknet Cool announcement, but respect to whoever designed your logo

privacy pools is now live on @Starknet deposit ETH, USDC, or STRK, withdraw to a fresh address, and use any app on starknet privately without revealing your onchain history or public identity starknet is the place for privacy

👉 Mastering Cairo Iterators ericnordelo.medium.com/mastering-cair…

Most Cairo devs still write while loops. But iterators in Cairo are not just cleaner, they’re cheaper. In benchmarks, iterator-based code is 20–30% more gas-efficient than equivalent loops. I wrote a full deep dive with examples, patterns, and benchmarks👇

Building on @Starknet keeps getting easier 🛠️ Cairo Contracts v3.0.0 is here - fully audited 🔒 Key updates: • ERC6372 clock standard for governance voting mechanisms • AccessControlDefaultAdminRules for enhanced admin management • Simplified ERC4626 fee configuration

@paradex true if big

Privacy Perps 🔒 Privacy Perps are now live on Mainnet. Effective today, position and account state on Paradex is private to the account holder (via authenticated RPC) and the operator (Paradex), delivering CEX-like privacy but with self-custody. This privacy is is end-to-end: orders, positions and trades are private on the Paradex Cloud, Paradex Chain (L2) and on Ethereum (L1) where state commitments are posted. Prior to today’s upgrade (Paradex Chain v0.14.1), there were two broad ways to observe or reconstruct account/chain state: 1️⃣ querying off-chain services that index Paradex 2️⃣ reconstructing aggregated balances from the L1 data availability v0.14.1 closes both paths for unauthenticated observers while preserving an escape hatch on Ethereum. Architecture L2: Private reads via authenticated RPC Privacy on the L2 is guaranteed through a custom RPC configuration. RPC nodes mask position and account fields by default, and only return private data to the owner after authenticating the request via a valid signature. This addresses (1) the historical “off-chain indexing” visibility path. Bridge partners are only granted selective access to the transaction data (account state remains private) required by their smart contracts to process deposits and withdrawals. L1: Encrypted state diffs with ZK verification When the sequencer posts updates to Ethereum, it will now encrypt the state diff and include it in the ZK proof posted to the L1. This specifically addresses (2) the historical “reconstruct state from L1 DA” path. Starting in v0.14.1, the blob data that carries the serialized squashed state diff is encrypted before being posted to Ethereum, while the ZK proof still validates both the correctness of the state transition and the correctness of the encryption/commitment to the encrypted diff. Decryption keys are held by the Paradex Privacy Council, and only council members can decrypt the L1 state. A more technical write-up, including encryption details, is available here: docs.paradex.trade/trading/privacy Why a Privacy Council Paradex continues to submit state to Ethereum to preserve the possibility of an escape hatch in the event Paradex becomes inoperable. The council model balances that escape hatch with privacy: under normal operation diffs remain encrypted on L1, and if an escape hatch is required, the council can decrypt and publish the necessary state to enable recovery. The current members of the Privacy Council are the @ParadexFNDN, @paradex, and @karnotxyz. As Paradex scales, the committee will be expanded for even stronger decentralization. Trust assumptions At the RPC layer, trust assumptions do not change: Paradex remains the operator of Paradex Chain RPC nodes. At the Ethereum layer, trust shifts to the Privacy Committee for the ability to decrypt state diffs during an escape-hatch scenario. TL;DR + Private by default on L2 + encrypted diffs on L1, with the possibility of an escape hatch preserved via council-controlled decryption. + Only the account owner and operator can view their trading state, providing a CEX-equivalent privacy experience with self-custody. Paradexio

@peter_szilagyi @protolambda Is space, just leave a window open

@protolambda How would cooling be done though?

When will we see the first bitcoin mining in space? If reduced energy cost (solar), and battery avoidance (24/7 source) are the main value proposition of computing in space, this could benefit btc mining too.

@Lyskey Not saying I dont agree but define success

Internet is truly incredible. It gives you access to every business, market, and piece of knowledge, giving opportunities to success for anyone, anywhere. Yet most people use it to watch TikTok dances, live through other people’s lives on Instagram, and just kill their attention span. Natural selection at a planetary scale.

@PatrickAlphaC What is this "connected via" nonsense? What is the argument behind making not only the country you are based in public, but also the one you currently are in while traveling?

This is a bad feature. Privacy is normal.

@hanni_abu I mean... I have to say it, it is great meme 😂😂😂

Starknet: Why doesn't the Ethereum community support us? Starknet booth at devconnect:

@MentorNotPseudo @EliBenSasson @mert When barber?

Bald following us everywhere @EliBenSasson @mert from avnu hq

@OhadBarta When you mention "applications that SHARP proves", like A and B, what does application mean in this context? Is Starknet one by itself? Are app chains different ones?

Today, when talking with a really-smart-and-amazing ecosystem player, I realized that SHARP architecutre suffers from gap of understanding. Not as mathematically sofisticated as the prover itself. Not as user facing as Starknet. Still, SHARP is the amazing and complicated beast that invokes the prover, get proofs, aggregates them, and make Starknet the most efficient on-chain rollup. So it deserves a deep dive. I'll start by answering the question of: "what happens when a Starknet block gets to sharp? Where S-Two is invoked and where Stone?". But feel free to add more questions. I'll shed light over anything you want Step 1 - SHARP gets the job When SHARP gets a job, its start with validating the job can be proven by re-execute the Cairo and inspect its size. Then, it assingns each job to its "applicative tree". Why applicative? Because there is one per application SHARP proves Why tree? Because of the proof structure The goal - condensing many blocks (block 1 with state transition 1-->2, block 2 with state transition 2-->3,etc) to one claim over correct execution of many blocks (with state update 1-->n). How its done in practice? After SHARP produces a proof for block "1" and block "2", its then proofs a claim of "I ran a successful validation" of blocks 1 and 2, generating one proof for the claim "1-->3". This process continues until you have a single root with all the jobs to prove that proves the state "1-->n" THIS PART HAPPENS WITH S-Two, and its much faster than it was in the Stone day! Roughly one hour faster Step 2 - SHARP aggregates many applicative trees into one "train" Ok, so each applicative tree now has its own proof and its own state update, but we still don't to send one proof to Ethereum per, say, 100 SN blocks. We need to be cheaper! So this is the part where several applicative trees are combined into one proof. Very similarly to the recusrsion inside the applicative tree, SHARP is saying something like: "here is a proof that application A moved from 1-->n and application B from 1'-->n'". THIS PART ALSO HAPPENS WITH S-Two Step 3 - SHARP wraps everything with a Stone proof Wait, what? But you have S-Two. True, but S-Two's verfiier was not implemented in Solidity yet, and the fact is we have several optimizations to do there before it would make sense to start the project of writing this code in Solidity. We are now working on these optimizations! Fear not. The day of S-Two on Ethereum will come (though not really, really soon). That's it, the first SHARP deep dive came to an end. Some things here are intentionally simplified . And I only had like 20 minutes before my wife would kill me for being late home, but couldn't stop myself. So if something doesn't click - its probably me, and not you. Ask questions so I know you arrived to the end, or throw an egg on me so I'll know to get back to Starknet-focused content. That's also valid!

OpenZeppelin Solidity Contracts v5.5 is here 💥 Gas optimizations, native passkey authentication, formal verification support, and cross-chain interoperability—all designed to help you build secure, efficient smart contracts. Here's what's new 🧵

@itamarl add crypto neobanking before it was cool somewhere in your profiles

This podcast was filmed in 2020 5 years before the "crypto neobank" meta