Announcing SpaceKit: The First-Ever Swift Smart Contract Framework 🙌 After 10 months of development, I’m excited to introduce SpaceKit, the first smart contract framework built for Swift 🔥 This marks a major milestone for @MultiversX and every Lightspeed Chain, making it the first blockchain ecosystem where Swift developers can build decentralized applications What I’ll cover in this post 1) SpaceKit’s Vision: A framework designed with developers in mind 2) Current Features: What’s available today 3) The SwiftVM: A Swift-based replica of SpaceVM for testing 4) Documentation & AI Assistance: Interactive tutorials and a GPT-trained assistant 5) The Future of SpaceKit: What’s next for the framework A TL;DR is included at the end, but let’s go through the details 1) SpaceKit’s Vision: A Developer-First Smart Contract Framework SpaceKit is designed to provide a smooth developer experience, leveraging Swift’s powerful and intuitive syntax while abstracting away unnecessary complexity - Cross-platform support: Works on macOS, Linux, and Windows (via WSL) - Flexible development environment: Use Xcode or VSCode - Fully compatible with MultiversX and all future Lightspeed Chains - A Unified Development Stack The long-term goal of SpaceKit is to enable developers to build smart contracts, backends, and frontends within the same Swift-based ecosystem: - Define contract endpoints - Automatically generate interaction functions - Develop entire decentralized applications using a single codebase Currently, SpaceKit is in beta and focused on smart contract development, supporting the most important opcodes. You can already build anything from NFT collections to DeFi protocols, with more capabilities coming soon. 2) Current Features: What’s available today SpaceKit is already packed with useful tools. Here’s a look at what’s possible: - High-Level Type System: Work with Buffer, Vector, BigUint, and more instead of raw byte operations - Intuitive Storage Management: Use @​Storage annotations or custom storage mappers like WhitelistMapper, VecMapper, and SetMapper - Blockchain & Transaction Data Access: Retrieve timestamps, payments, caller addresses, and other contextual information - Flexible Coding Style: Write contracts using global functions, OOP-style structs, or simple @​Codable annotated structs - Simplified Contract Interactions: Deploy, upgrade, and call other contracts with minimal effort thanks to the @​Proxy annotation - No low-level complexity: SpaceKit abstracts away low-level intricacies, automatically determining whether the code is being compiled for the SpaceVM or SwiftVM. This eliminates the need for Self::Api and other generic constraints, allowing you to focus on building business logic without dealing with pointers or WebAssembly-specific details - Async Calls with Callbacks: Mark functions with @​Callback to handle asynchronous contract interactions - Full ESDT Support: Issue, mint, burn, and manage fungible, non-fungible, and semi-fungible tokens - Secure Randomness – Generate random numbers using SpaceVM’s built-in random features - Event Emission: Use @​Event to automatically generate event methods - Command Line Interface: Initialize projects, compile to WebAssembly, and generate ABIs - Rust Compatibility: Swift smart contracts can interact seamlessly with Rust contracts - Cross-Platform Support: Fully functional on macOS, Linux, and Windows - Code Editor Support: Full autocompletion and syntax highlighting, even for generated code - Compatible with Any SpaceVM-Powered Blockchain These features enable developers to build almost any type of smart contract 3) The SwiftVM: A Swift-Based Testing Environment Testing is a critical part of smart contract development, and SpaceKit simplifies it with SwiftVM, a Swift-based replica of SpaceVM. This allows developers to: - Run contract endpoints and check results - Simulate failures and analyze error reasons - Set up wallet balances and simulate transactions - Define which address is calling the contract - Test contract-to-contract interactions, both synchronous and asynchronous - Ensure callbacks are working, even when the asynchronous call fails - Issue and manage ESDT tokens within tests through a replica of the GoVM’s ESDT System smart contract - Use Swift’s built-in debugger and breakpoints for in-depth testing - This eliminates the need for external testing tools, allowing developers to test their contracts natively in Swift 4) Documentation & AI Assistance To make onboarding easy, SpaceKit includes: - An 11-Chapter Interactive Tutorial: Covers installation, contract development, ESDT management, and more - 22+ Example Smart Contracts: Ranging from simple adders to fully functional multisigs - A Repository with 568+ Tests: Demonstrates real-world testing scenarios - Custom GPT Assistant: Provides instant answers to development questions More documentation and learning resources will be added over time 5) The Future of SpaceKit The roadmap includes several key milestones: Short-Term: Incorporate developer feedback, refine documentation, and expand learning resources Mid-Term: Improve the CLI and SwiftVM, reducing reliance on external tools Long-Term: Develop a Swift-based backend framework, similar to NovaX, for seamless contract interactions The ultimate vision is a fully integrated Swift ecosystem where developers can build smart contracts, backends, and even iOS/macOS frontends from the same codebase TL;DR - SpaceKit: The first Swift smart contract framework - Developer-Focused: Works across platforms, with an intuitive API and minimal complexity - Comprehensive Feature Set: Easy storage, event emission, async calls, and more - SwiftVM: Test contracts directly in Swift with real debugging tools - Extensive Documentation & AI Support: Tutorials, example contracts, and an AI assistant - Future Plans: A full backend + frontend Swift ecosystem for dApp development SpaceKit is here. Ready to start building? 🔥 - SpaceKit repository: github.com/gfusee/SpaceKit - Interactive tutorials: gfusee.github.io/SpaceKit/tu… - GPT assistant: chatgpt.com/g/g-6793b3de291c…

this gave us the experience to now be able to review efficiently what AI produces I don't code by hand now, yet I'm so grateful that I had to do it in the past and learnt SWE that way

friendly reminder to do some kind of sport every day helps a lot when you’re in the trenches

"At this pace, AI will replace <insert any role>" - anyone making money from AI market growth

I updated my @SuiNetwork + Walrus + Seal localnet docker images to use the latest version 1.81.0 of Sui They are mainly for my own purposes, I put the setup open-source in case of someone needs it as well It has a few e2e tests ensuring the 3 components act well together The update is represented by the tag 0.0.23: github.com/gfusee/sui-localn…

we forget how much of a game changer better ai tooling can be sometimes even more than upgrading the models themselves dots is a great example

almost every component needed to abstract blockchain away from users already exists on @SuiNetwork but there’s still one missing piece think of the sui ecosystem as a decentralized cloud stack → sui to secure and orchestrate apps and assets → walrus to store data → seal to keep that data private to authorized users → nautilus to handle complex computations → deepbook to provide liquidity infrastructure → zklogin to remove the need for users to manage private keys put all of this together, and you can build applications where users barely need to know a blockchain is involved... ...almost there’s still one thing missing, can you guess what it is? let me give you a few hints, imagine you’re building a product backed by a web3 infrastructure and want the blockchain to be 100% invisible to your users this is our goal at @PrtclMediaLabs there’s still one thing standing in your way: → almost every web2 product has one → when you hear “SaaS”, it’s probably one of the first things that comes to mind you got it: web2-native billing subscriptions and one-shot payments through credit and debit cards, not crypto for a builder, the real problem isn’t just accepting a payment it’s being able to charge users automatically when the next payment is due because a subscription where users have to manually come back and pay every month isn’t really a subscription every extra action is another opportunity for the user to drop and that friction is enough to kill a business built on recurring revenue but building the payment infrastructure yourself isn’t really an option either because payments are an entire business of their own: compliance, refunds, disputes, claims, fraud, payment methods, currencies none of this is what a builder creating an app on sui should have to spend time on and this is not something other blockchain ecosystems have made a first-class part of their stack either they’ve built ways to move tokens, some have even built ways to automate recurring token payments but that’s not the same as giving builders web2-native billing -> a user should be able to enter their card once and forget about it -> the builder should be able to charge them once or every month and neither side should have to care about wallets, tokens, on-ramps or what happens underneath this is where sui has a card to play it already has most of the infrastructure needed to make blockchain invisible web2-native billing could be one of the last pieces needed to make the stack feel complete if we really want blockchain to disappear from the user experience, payments have to disappear with it

all those animations created with opus 5.5 are impressive but they all look the same I can’t wait to see what happens when motion designers bring their own taste to these models

y'all wasting your energy chasing the latest AI model stick to a model you like and use that energy to build things that actually have value

the fix is to let JavaScript die and replace it by Rust

a case study of why @SuiNetwork’s object model is so powerful when building @PrtclMediaLabs we ran into a simple problem: writers care about writing, not blockchain we needed a system where they could configure everything on-chain and remain fully self-custodial while still allowing us to update their settings on their behalf this enables features that are essential for a web2-like experience: automatic royalty off-ramps, managed encryption key rotation, … the problem? the value of web3 is transparency and censorship resistance giving one entity (us) these superpowers goes against the vision we have at first it seemed contradictory we wanted a system where we couldn’t control the database, while still being able to act on behalf of users when needed so we took advantage of sui’s object model to make both possible that took 3 steps: 1) building a fine-grained permission system the solution is based on something standard in the web2 industry each exposed Move function has an associated permission they are listed in an enum called MemberPermission, with permissions such as: - UpdatePublicationPricing - ClaimRevenues - ArchiveArticle 2) defining publication members the owner of a publication automatically has every permission in MemberPermission but a publication can have multiple writers to invite one, the owner mints a Sui object called Member, assigns it a set of permissions and sends it to the writer’s wallet holding this object gives the writer access to exactly the actions they were granted 3) allowing Protocol to mint Members transparently let’s say a publication owner isn’t tech savvy and wants to delegate the management of article encryption keys to Protocol they can optionally give Protocol the ability to mint special Member objects, called ProtocolMember but these objects are: -> scoped to only the permissions authorized by the owner -> valid for only a few minutes think of it as an on-chain agreement the owner tells Protocol: “I trust you to give wallets you control these exact permissions on my behalf” and Protocol answers: “we built the infrastructure so it’s technically impossible for us to use that power without you knowing” now let’s get back to encryption key rotation rotating a key requires decrypting the articles, encrypting them again and publishing the new keys the first action happens off-chain through Seal, and anyone authorized on-chain by the owner can do it but remember: the owner didn’t give Protocol permission to decrypt articles they gave Protocol permission to give one of its wallets permission to decrypt articles, through ProtocolMember objects that distinction is important for Protocol to decrypt an article, it first has to mint a ProtocolMember object: -> for a specific publication -> with a specific permission -> for a specific amount of time in other words, Protocol can decrypt articles when the owner has authorized it to do so but it cannot exercise that power without first saying publicly, on-chain and in a fully auditable way: “hey, I’m using my superpower for the next 5 minutes” this is where Sui’s object model shines because permissions themselves are objects, auditing this becomes extremely natural a researcher can look through the blockchain history for ProtocolMember objects and see when delegated access was created, what it allowed and how long it was valid for on an account-based model, the same information can be spread across contract state, events and transactions with Sui, the capability itself is a first-class object the permission system becomes the audit trail

after updating and monitoring everything I confirm that bumping your @SuiNetwork deps to 1.80.0 works I highly recommend you to update as soon as possible, as even if mysten mitigates this, there are still risks that your critical on-chain components stop working suddenly this is especially true for services parsing on-chain transaction data

yesterday the sui community lost big, we'll miss you frank 🕊️

a journalism platform's job is to give its customers the best possible access to truthful, transparent, and independent information if blockchain helps you do that, use it protecting the way you’ve always done things is a terrible reason to give customers less than what’s possible

building in the @SuiNetwork ecosystem is fantastic, until you need to test we don’t have anything close to what the EVM ecosystem has with Hardhat, and many of the simulation tools we need simply don’t exist yet the problem gets even worse for projects like @PrtclMediaLabs that use the entire Sui stack: Walrus, Seal, DeepBook, zkLogin, and more so early in our building process, we made a decision: treat our testing environment as more valuable than the code it tests to make that possible, we built our own Sui stack simulator in-house -> an entire program embedding the MoveVM and Sui execution layer, extracted from the Sui repo and rewired around our internal needs there’s no consensus layer, and we have admin-level access to inject objects and packages from scratch, letting us recreate real-world scenarios in no time what does this mean? real PTBs. real transaction execution. full control: - time can be controlled - checkpoints can be advanced at will - transactions can be executed without having the sender’s private key - Walrus uploads can be delayed to simulate real-world latency simulated Seal servers connect in-process directly to the Sui simulator, so integration tests run quickly and a complete environment takes just 20-30ms to set up there is literally nothing on-chain we cannot test or replicate the simulator also exposes replicas of the real gRPC endpoints used by the official Sui SDKs, making them believe they’re communicating with an actual Sui network even better: the same simulator is used for local manual testing with one command, devs can spin up a fresh environment on their own machine

the past few months made me realize again something about @SuiNetwork: it has never been that easier to build an app that barely feels like web3 at @PrtclMediaLabs we put together Sui, Seal, Walrus, DeepBook and zkLogin to build and deploy on mainnet an entire and decentralized journalism platform that journey went through many highs, a few lows, and hundreds of technical things I want to share about what we created I’ve never been more convinced that blockchains, as infrastructure, can change the world

gm gm, what did I miss during the summer????

Hellooooooo all, at @PrtclMediaLabs, we're in the final stages of building and we're looking for fiat ↔ crypto payment providers that: - are compatible with @SuiNetwork - support subscriptions and recurring payments If you have any contacts, my DMs are wide open! 🙏

ran multiple build at the same time on a macbook pro m5 max, it made the computer crash lmao