Those Who Swift - Issue 252
Your weekly dose of Swift!
Weekly note ✏️
This week, OpenAI finally unveiled its standalone coding tool: the Codex app for macOS. Feature-wise, it brings OpenAI into direct parity with existing competitors, putting the market into a truly competitive state. Pricing and core functionality are now largely comparable across major players, which means the real differentiators will be the depth of third-party integrations and the quality of the underlying models.
So what does this mean for Apple? Last year, Apple acquired the AlexSidebar tool, positioning it as one of the few assistants integrated directly into Xcode. Yet most advanced iOS development workflows involving MCPs or agents are still happening outside Xcode, usually via CLI tools or IDEs like Cursor or VS Code. This gap has increasingly pushed Xcode into the shadow of being perceived as outdated. And latest non-WWDC video in our Video section is showing it.
Apple has started adding features outside the WWDC cycle, likely to signal that active work is happening. We’re also seeing organizational changes: a major shift in AI leadership, with renewed focus on Siri powered by Gemini. It’s unclear how much betting on a personal assistant will help Apple compete in the broader developer tooling race, but one thing is certain—change needs to happen quickly.
On the design side, there’s movement too. The creator behind Halide, Kino, and Orion joining Apple’s Design Team. Expectations are high. Still, in today’s landscape, announcing major software updates just once a year feels increasingly slow—especially when months pass between announcement, RC, and real availability.
Connect with the "Those Who Swift" team - Justas Markus & Anton Gubarenko 👋
Sponsor 🤝
Forget about Ruby and Fastlane installation issues!
Discover Codemagic CLI tools — the free, open-source Fastlane alternative for automating iOS builds, code signing and publishing.
Swift Around the Web 🌐
How Apple Hooks Fifty Thousand Methods
This video by Bryce Bostwick explores method swizzling in Objective-C and Swift — how it works, common use cases, and the risks it introduces (like unexpected behavior and maintenance challenges). It shows patterns for applying swizzling safely and when it’s better to avoid it entirely in modern code.
Tiered Caching in Swift
Kyle Browning explains how to build a tiered caching system in Swift that layers fast in-memory caches with slower disk or network caches: improving performance and reliability by returning the best available result while keeping data fresh and fallback paths in place.
Objectively Better, Observably Trickier
This post from Danny Bolella explores subtleties in observable state handling in SwiftUI, showing how different observation approaches affect view invalidation and performance. It highlights practical examples where small design choices lead to unexpected redraws or missed updates, helping developers make more predictable, efficient state models.
Coding 👨💻
The Secret to Buttery Smooth SwiftUI
This article explains key SwiftUI performance principles and shows how to reduce unnecessary view updates using techniques: with false and optimal solution examples.
Designing Swift Errors for an SDK
Guide covers how to design error types in Swift SDKs for clarity and usability. It discusses creating meaningful, structured Error cases, using associated data effectively, and providing developers with clear debugging information without leaking implementation details.
Design 🎨
Branding With AI Superpowers for Indie Creators
This article explores how indie app makers can use AI tools to strengthen their branding by generating consistent visuals, tone, and messaging. It highlights practical ways to combine creativity with AI assistance to craft memorable identities without a big design team.
Print Design Mindset to App Design Mindset
This essay from Christopher Trott reflects on adopting a print design mindset in digital work: focusing on clarity, typography, and thoughtful layout as if preparing content for physical media. The author argues that designing with print principles can improve readability and visual hierarchy in apps and interfaces.
Other cool stuff 🧰
How to Make Money With Newsletter?
The owner of Those Who Swift and Indie App Devs newsletters is sharing tips on how to make money with newsletters. From choosing the right platform and growing your audience, to finding sponsors and even buying an existing newsletter if you don’t want to start from scratch.
SwiftUI Weather App Copy Cat With WeatherKit
Itsuki walks through building a simple but powerful SwiftUI weather app using Apple’s WeatherKit, showing how to fetch and display current, hourly, and daily forecasts tied to a location. He also highlights setup steps, data expiration handling, and tips for presenting weather info cleanly in your app.
From Objective-C to Swift 6: What We Gained
This guide walks through the Swift evolution process, explaining how proposals are drafted, reviewed, and accepted (and how changes to the language are shaped by community feedback and core team decisions). It’s a clear roadmap for anyone interested in contributing to or understanding Swift’s ongoing development.
AI 🤖
Introducing the Codex App
OpenAI has launched the Codex app, a new tool designed to help developers write, edit, and understand code more efficiently using AI: blending intelligent completions with contextual assistance directly in your coding workflow.
Supporting Markdown Search for LLMs
Learn how to make your blog easier for AI models to search by serving Markdown versions of your content, which use far fewer tokens than full HTML and fit better into LLM context windows. It shows how to expose .md URLs and redirect LLM requests so tools like Claude Code and Codex can retrieve cleaner, lightweight content more efficiently.
Claude Code Plugins for iOS Teams, Automation Agents and Guardrails
Wesley Matlock dives into how Claude Code plugins can enhance iOS development workflows by automating routine tasks, enforcing guardrails, and coordinating multi-step processes with AI agents tailored to team needs. He explores practical plugin patterns and examples that help teams build faster, safer, and more consistent apps.
Tutorials 📒
A WhatsApp-Style Top Banner for iOS Using UIWindow
Artem Mirzabekian kindly shows how to build a WhatsApp-style top notification banner in SwiftUI by presenting it in a separate UIWindow. The approach avoids layout conflicts, supports smooth animations, and keeps the banner independent from your main view hierarchy.
Video 🎥
Meet agentic coding in Xcode
This Apple Tech Talk shows how Xcode 26.3 lets coding agents like OpenAI Codex and Claude Agent collaborate on multi-step tasks (building projects, running tests, and querying Apple docs) using the Model Context Protocol. You’ll learn how this integration streamlines agent-driven development workflows directly inside Xcode.
Mastering DatePicker, MultiDatePicker, and ColorPicker in SwiftUI
Stewart Lynch walks through SwiftUI’s picker essentials, covering DatePicker, MultiDatePicker, and ColorPicker with real-world patterns and pitfalls. You’ll learn how to handle ranges, styles, optional dates, multi-date selection, color persistence, and when to bridge to UIKit for full control.
Thanks for reading Those Who Swift! Subscribe for free to receive new posts.
Personally, I prefer Apple products to evolve more slowly. The Photos app changes often, and this introduces issues for non–tech-savvy users. AI in every app? Sorry, but UX is already changing daily, and AI tools are evolving just as fast.
I don’t want tools that exist just to push something new. Who would want to spend three months writing a book about AI when, three months later, half the information is outdated?
Rapid development belongs on other platforms. Apple platforms should evolve more slowly, especially during a volatile technological shift like the current AI wave.
Solid analysis on Xcode's positioning problem. The point about Apple needing to move faster than the yearly WWDC cycle is critical, especially when competetors like Cursor are shipping weekly updates. I've noticed this friction myself when setting up AI tooling for Swift projects. Integration depth matters way more than raw model quality atthis point.