# PeopleInSpace ![kotlin-version](https://img.shields.io/badge/kotlin-2.4.0-blue?logo=kotlin) **Kotlin Multiplatform** project with SwiftUI, Jetpack Compose, Compose for Wear OS, Compose for Desktop and Compose for Web clients along with Ktor backend. Currently running on * Android (Jetpack Compose) * Android App Widget (Compose based Glance API - contributed by https://github.com/yschimke) * Wear OS (Compose for Wear OS - primarily developed by https://github.com/yschimke) * iOS (SwiftUI) * Desktop (Compose for Desktop) * Web (Compose for Web - Wasm based) * Windows (WinUI 3) * JVM (small Ktor back end service + `Main.kt` in `common` module) * MCP server (using same shared KMP code) The people data comes from [The Space Devs API](https://thespacedevs.com/llapi) (names, bios and images of the people currently in space) and the position of the International Space Station from the [Where the ISS at? API](https://wheretheiss.at/w/developer), both served through this project's own small Ktor backend (see `backend` module below). The project is included as sample in the official [Kotlin Multiplatform docs](https://kotlinlang.org/docs/multiplatform-samples.html) and also the [Google Dev Library](https://devlibrary.withgoogle.com/products/android) ### Architecture All of the clients are thin UI layers over the same `common` module: the shared view models expose `StateFlow`s of UI state, backed by a single repository that treats the local SQLDelight database as the source of truth and refreshes it from the network. ```mermaid flowchart TB subgraph clients["Clients"] direction LR android["app
Android · Jetpack Compose
+ Glance / Remote Compose widgets"] wear["wearApp
Wear OS · Compose + Tile"] ios["PeopleInSpaceSwiftUI
iOS · SwiftUI (Swift Export)"] desktop["compose-desktop
JVM · Compose for Desktop"] web["compose-web
Kotlin/Wasm · Compose"] winui["windows/WinUiApp
.NET · WinUI 3"] mcp["mcp-server
JVM · Kotlin MCP SDK"] end subgraph common["common (Kotlin Multiplatform)"] direction TB cmpui["Compose Multiplatform UI
PersonList / ISSPosition / ISSMapView
(expect/actual map per platform)"] vm["View models
PersonListViewModel · ISSPositionViewModel
StateFlow<UiState>"] winclient["PeopleInSpaceClient
mingwX64, exported via NuGet
(no Koin / AndroidX)"] repo["PeopleInSpaceRepository
offline-first: DB is source of truth,
ISS position polled every 10s"] db[("SQLDelight
PeopleInSpaceDatabase")] api["Ktor client APIs
PeopleInSpaceApi · AstroviewerApi"] koin{{"Koin DI
(annotations + compiler plugin)"}} cmpui --> vm vm --> repo winclient --> repo repo --> db repo --> api koin -.-> repo end subgraph remote["Remote"] direction LR backend["backend
Ktor / Netty on App Engine
/astros.json"] spacedevs["The Space Devs API
names, bios, images"] wheretheiss["wheretheiss.at
current ISS position"] astroviewer["astroviewer.net
predicted ISS orbit"] end android --> cmpui wear --> repo ios --> cmpui desktop --> cmpui web --> cmpui winui --> winclient mcp --> repo api --> backend api --> wheretheiss api --> astroviewer backend --> spacedevs classDef client fill:#e3f2fd,stroke:#1565c0,color:#0d1b2a classDef shared fill:#ede7f6,stroke:#5e35b1,color:#0d1b2a classDef data fill:#e8f5e9,stroke:#2e7d32,color:#0d1b2a classDef service fill:#fff3e0,stroke:#ef6c00,color:#0d1b2a class android,wear,ios,desktop,web,winui,mcp client class cmpui,vm,winclient,koin shared class repo,db,api data class backend,spacedevs,wheretheiss,astroviewer service ``` Notes on a few of the edges above: * The Wear OS client and the MCP server talk to the repository directly (Wear has its own Wear-specific view models, the MCP server just reads the people list). * The Windows client goes through `PeopleInSpaceClient`, a self-contained `mingwX64` entry point that owns its own Ktor engine, SQLite driver and coroutine scope rather than using Koin. * Only the people list goes through this project's own Ktor backend; the ISS position and predicted orbit are fetched from their services directly by the shared Ktor client code. ### Module overview | Module | Description | |---|---| | `common` | Shared KMP code (Ktor, SQLDelight, Koin, view models), shared Compose Multiplatform UI, and the Windows NuGet API | | `app` | Android client (Jetpack Compose), including Glance app widget | | `wearApp` | Wear OS client (Compose for Wear OS) | | `PeopleInSpaceSwiftUI` | iOS client (SwiftUI) | | `compose-desktop` | Desktop client (Compose for Desktop) | | `compose-web` | Web client (Compose for Web, Kotlin/Wasm) | | `windows/WinUiApp` | Windows client (WinUI 3) | | `backend` | Ktor server providing the people/ISS data (deployable to Google App Engine) | | `mcp-server` | Model Context Protocol server exposing the shared KMP code | ### Building and running Requirements: JDK 17, a recent version of Android Studio (for the Android/Wear clients) and Xcode (for the iOS client). * **Android**: run the `app` configuration from Android Studio (or `./gradlew :app:installDebug`) * **Wear OS**: run the `wearApp` configuration from Android Studio (or `./gradlew :wearApp:installDebug`) * **iOS**: open `PeopleInSpaceSwiftUI` in Xcode and run from there * **Desktop**: `./gradlew :compose-desktop:run` * **Web (Wasm)**: `./gradlew :compose-web:wasmJsBrowserDevelopmentRun` * **Windows (.NET)**: see the [WinUI 3 client guide](windows/README.md) for prerequisites and build, test, and run instructions. * **Backend**: `./gradlew :backend:run` (or run `Server.kt` directly from Android Studio). After doing that you should then for example be able to open `http://localhost:9090/astros_local.json` in a browser. Tests can be run with `./gradlew :common:jvmTest`, and there's also a [Maestro](https://maestro.mobile.dev/) UI test flow shared by the Android and iOS clients, run by CI on an emulator and a simulator (`maestro test -e APP_ID=dev.johnoreilly.peopleinspace maestro/PeopleInSpace.flow`). ### Backend deployment Have tested this out in Google App Engine deployment. Using shadowJar plugin to create an "uber" jar and then deploying it as shown below. Should be possible to deploy this jar to other services as well. ``` ./gradlew :backend:shadowJar gcloud app deploy backend/build/libs/backend-all.jar --appyaml=backend/src/jvmMain/appengine/app.yaml ``` ### Screenshots **iOS (SwiftUI + CMP)**
Screenshot 2025-08-10 at 22 15 21 **Android (Jetpack Compose + CMP)**
Screenshot 2025-08-10 at 22 15 29 **Wear OS (Wear Compose)**
Wear OS person list Wear OS list with ISS map edge button Wear OS ISS position map **Compose for Desktop**
Screenshot 2025-08-10 at 22 51 15 **Compose for Web (Wasm based)**
Screenshot 2025-08-10 at 22 58 45 **MCP** The `mcp-server` module uses the [Kotlin MCP SDK](https://github.com/modelcontextprotocol/kotlin-sdk) to expose an MCP tools endpoint (returning list of people in space) that can for example be plugged in to Claude Desktop as shown below. That module uses same KMP shared code (that uses for example Ktor, SQLDelight and Koin) ![Gry_C3FXkAAxVvN](https://github.com/user-attachments/assets/74c210b9-9a0a-4de8-8845-81380f11e4a5) To integrate the MCP server with Claude Desktop you need to firstly run gradle `shadowJar` task and then select "Edit Config" under Developer Settings and add something like the following (update with your path) ``` { "mcpServers": { "kotlin-peopleinspace": { "command": "java", "args": [ "-jar", "/mcp-server/build/libs/serverAll.jar", "--stdio" ] } } } ``` ### Related posts * [Minimal Kotlin Multiplatform project using Compose and SwiftUI](https://johnoreilly.dev/posts/minimal-kotlin-platform-compose-swiftui/) * [Adding some Storage (to) Space](https://johnoreilly.dev/posts/adding-sqldelight-to-peopleinspace/) * [Kotlin Multiplatform running on macOS](https://johnoreilly.dev/posts/kotlinmultiplatform-macos/) * [PeopleInSpace hits the web with Kotlin/JS and React](https://johnoreilly.dev/posts/peopleinspace-kotlinjs/) * [Using Koin in a Kotlin Multiplatform Project](https://johnoreilly.dev/posts/kotlinmultiplatform-koin/) * [Jetpack Compose for the Desktop!](https://johnoreilly.dev/posts/jetpack-compose-desktop-copy/) * [Comparing use of LiveData and StateFlow in a Jetpack Compose project](https://johnoreilly.dev/posts/jetpack-compose-stateflow-livedata/) * [Wrapping Kotlin Flow with Swift Combine Publisher in a Kotlin Multiplatform project](https://johnoreilly.dev/posts/kotlinmultiplatform-swift-combine_publisher-flow/) * [Using Swift Packages in a Kotlin Multiplatform project](https://johnoreilly.dev/posts/kotlinmultiplatform-swift-package/) * [Using Swift's new async/await when invoking Kotlin Multiplatform code](https://johnoreilly.dev/posts/swift_async_await_kotlin_coroutines/) * [Exploring new AWS SDK for Kotlin](https://johnoreilly.dev/posts/aws-sdk-kotlin/) * [Creating a Swift command line app that consumes Kotlin Multiplatform code](https://johnoreilly.dev/posts/swift-command-line-kotlin-multiplatform/) * [Exploring New Worlds of UI sharing possibilities in PeopleInSpace using Compose Multiplatform](https://johnoreilly.dev/posts/exploring-compose_multiplatform_sharing_ios/) * [Moving from SKIE to Swift export in a Kotlin Multiplatform project](https://johnoreilly.dev/posts/peopleinspace-swift-export/) ### Related samples Note that this repository very much errs on the side of minimalism to help more clearly illustrate key moving parts of a Kotlin Multiplatform project and also to hopefully help someone just starting to explore KMP to get up and running for first time (and is of course primarily focused on use of Jetpack Compose and SwiftUI). If you're at the stage of moving beyond this then I'd definitely recommend checking out [KaMPKit](https://github.com/touchlab/KaMPKit) from Touchlab. I also have the following samples that demonstrate the use of a variety of Kotlin Multiplatform libraries (and also use Jetpack Compose and SwiftUI). * GalwayBus (https://github.com/joreilly/GalwayBus) * Confetti (https://github.com/joreilly/Confetti) * BikeShare (https://github.com/joreilly/BikeShare) * FantasyPremierLeague (https://github.com/joreilly/FantasyPremierLeague) * ClimateTrace (https://github.com/joreilly/ClimateTraceKMP) * GeminiKMP (https://github.com/joreilly/GeminiKMP) * MortyComposeKMM (https://github.com/joreilly/MortyComposeKMM) * StarWars (https://github.com/joreilly/StarWars) * WordMasterKMP (https://github.com/joreilly/WordMasterKMP) * Chip-8 (https://github.com/joreilly/chip-8) ### Languages, libraries and tools used * [Kotlin](https://kotlinlang.org/) * [Kotlin Coroutines](https://kotlinlang.org/docs/reference/coroutines-overview.html) * [Kotlinx Serialization](https://github.com/Kotlin/kotlinx.serialization) * [Ktor client library](https://github.com/ktorio/ktor) * [Android Architecture Components](https://developer.android.com/topic/libraries/architecture/index.html) * [Koin](https://github.com/InsertKoinIO/koin) (using [Koin Annotations](https://insert-koin.io/docs/reference/koin-annotations/start) and the [Koin compiler plugin](https://insert-koin.io/docs/setup/compiler-plugin/)) * [SQLDelight](https://github.com/cashapp/sqldelight) * [Jetpack Compose](https://developer.android.com/jetpack/compose) * [Compose Multiplatform](https://www.jetbrains.com/compose-multiplatform/) * [Compose for Wear OS](https://developer.android.com/training/wearables/compose) * [SwiftUI](https://developer.apple.com/documentation/swiftui) * [Swift Export](https://kotlinlang.org/docs/native-swift-export.html) * [Coil](https://coil-kt.github.io/coil/) * [Kermit](https://kermit.touchlab.co/) * [Maestro](https://maestro.mobile.dev/)