Why did Go developers refuse to use segments when expanding the stack?

In the Go language, segments were abandoned when expanding the stack to simplify and improve the efficiency of goroutine stack management.

Instead of a segmented stack, Go uses a dynamically expandable stack that starts with a small size (usually a few kilobytes) and increases by copying data to a new, larger stack when necessary. This approach allows:

• To avoid the complexity of managing multiple segments and their connections.
• To simplify implementation and improve performance.
• To make the stack more flexible and adaptive to the actual needs of the goroutine.

Thus, abandoning segments is a compromise in favor of simplicity, efficiency, and ease of implementation of the scheduler.