Hands-On System Programming with Go

Workers

The previous example is a Google Maps client that uses a time.Ticker channel to limit the rate of the requests. The rate limit makes sense for an API key. Let's imagine that we have more API keys from different accounts, so we could potentially execute more requests.

A very typical concurrent approach is the workers pool. Here, you have a series of clients that can be picked up to process an input and different parts of the application can ask to use such clients, returning the clients back when they are done.

We can create more than one client that shares the same channels for both requests and responses, with requests being the coordinates and the results being the response from the service. Since the channel for responses is unique, ...