But for framework development, it’s a different story. Mixing Swift with Objective-C in a consumer app usually isn’t a big deal. Having a codebase that uses the same language There were a few factors that went into our decision to migrate to Swift. It was important for us to think about what we wanted our future SDK to look like and how we could start the evolution while also thinking about our company’s growth and scaling challenges.

We chose the tradeoff of a large up-front time investment to migrate in order to gain major long-term benefits.

Migrate completely to Swift and keep everything in one module (for now).Modularize our framework by breaking the pieces into separate modules and ship a more complicated dependency structure.Build a brand-new SDK that only supports StoreKit2, and continue to support the original SDK for people who want to stick with StoreKit1.If we wanted to support the new APIs while also supporting the original StoreKit, previous iOS versions, and Objective-C, we’d have to go through a significant effort to figure out how. Our project was already jumping through a few hoops to be distributed through CocoaPods, Swift Package Manager, and Carthage with our mixed-language support. At WWDC 2021, Apple announced a new version of the StoreKit framework (StoreKit 2), which would only support Swift and iOS 15. Our story, like so many indie development stories, begins with an announcement Apple made during WWDC. Let’s talk about our decision to migrate the preparation, development, and testing and finally, how we successfully launched it with minimal customer impact. We knew going into the migration that it wouldn’t be a trivial process, but we learned and experienced a lot of unexpected things along the way. We migrated our four-year-old Objective-C + Swift framework to pure Swift while retaining Objective-C support for our users with only the absolutely required API changes. When StoreKit2 was announced, we decided it was time to make the switch. Objective-C tooling and distribution were both stable and battle-tested, plus Jacob refused to learn Swift (and probably won’t ever have to - good job waiting that one out!).įour years later, Swift is mature and stable, and there’s a well-trodden path for distribution. At that time, Swift was still undergoing significant changes, and Swift SDK distribution was still not feasible in most cases. When we first started RevenueCat, Jacob wrote our first SDK.

Objective c to swift migration