Tempos limite, logging e URLs base

A real client needs configuration: timeouts so it never hangs, logging to debug, and a base URL to stay tidy.

You add behavior to the client using plugins, configured in the HttpClient builder block once for all requests.

Install the HttpTimeout plugin so a slow server fails fast instead of freezing your app forever.

The same plugin also sets a connectTimeoutMillis, capping how long Ktor waits to even reach the server.

The Logging plugin prints requests and responses, which is invaluable while debugging shared network code.

Pick a LogLevel like HEADERS or INFO to control noise, keeping verbose BODY logs out of release builds.

Use defaultRequest to define a base URL once, so each call only needs the path that follows it.

With a base URL set, your requests pass only the relative path, which keeps shared code clean and DRY.

defaultRequest can also attach shared headers, like a common API key, to every outgoing request automatically.

Set all this up a single time when you build the client, then reuse that one instance across the whole app.

Timeouts, logging and a base URL turn a toy client into a production-ready one that behaves well on both platforms. 🙂

Which plugin stops a request from hanging forever?

Install HttpTimeout, Logging and DefaultRequest once in the client builder to add safety, visibility and a clean base URL for every call.