Talks

Lots has been said about how to write testable code and about the importance of dependency management. But practically no one talks about how to write tests themselves. Why in some projects tests serve as documentation, and in others they only complicate maintenance? What to do if tests are being written, but the percent of code coverage is not increasing? How to turn a routine description of test cases into a creative process which is clear for all the team? The point of the talk is to answer these and many other questions, show the audience how to write readable and simple-to-maintain tests forming engineering culture.

Let's briefly recall what modularization is, why it's needed, and what Dependency Injection has to do with it. Let's think about how this principle, or rather the framework that implements it, helps with modularization.

Next we will discuss why Needle is the best choice among the existing open source DI frameworks. Let's also talk about Needle itself - how it is designed and what are the nuances of using it.

In the end, we'll return to how to use it in a multi-module application and how to circumvent the problems it causes. Let's not ignore the topic of testing either - let's talk about how to write tests on code that uses Needle.

The speaker share some nuances and features of iOS in-app subscriptions.

Consider what can be done based on SwiftUI Preview. Let's talk about automatic generation based on Previews: 

• Demo app
• Snapshot of tests
• Performance analysis views 
• Accessibility tests

We will try to generate all this completely automatically! In general, let's get the most out of SwiftUI Preview.

Which characteristics do you think have a stronger influence on the speed of building your project - the CPU frequency or the RAM frequency? The number of CPU cores or the amount of RAM? Does the speed of permanent memory affect the build speed? One day the speaker had such questions in his head, and he decided to find answers to them.

The speaker will explain what images are in computer memory, touch on the differences between UIImage, CGImage and CIImage, talk about Core Graphics and Core Image, and finish by running a PyTorch model through CoreML for simple classification.

We are building our own protocol for the p2p network, available to everyone on pure swift. Let's open up new possibilities for current application applications without headaches.

An exploration of the incremental compilation mechanisms found in today’s Android toolchain. The speaker will explain what an ABI is, discuss the «degrees» of incrementality and discuss how to develop an efficient build code base.

The talk is aimed at Android developers, but will also be useful for JVM-backend developers using Gradle. The iOS guys might be interested in broadening their horizons or getting inspired for their own research.