Coroutine scope and structured concurrency in Kotlin - Step-by-Step Execution

Concept Flow - Coroutine scope and structured concurrency

Start Coroutine Scope

↓

Launch Child Coroutines

↓

Child Coroutines Run Concurrently

↓

Wait for All Children to Complete

↓

Coroutine Scope Ends

↓

Continue with Rest of Program

This flow shows how a coroutine scope launches child coroutines, waits for all to finish, then ends, ensuring structured concurrency.

Execution Sample

Kotlin

fun main() = runBlocking {
  launch { delay(100); println("First") }
  launch { delay(50); println("Second") }
  println("Start")
}

This code runs two child coroutines inside a scope, printing messages with delays, showing structured concurrency.

Execution Table

Step	Action	Coroutine	Delay (ms)	Output	State
1	Enter runBlocking scope	main	0		Scope started, waiting for children
2	Launch first coroutine	first	100		First coroutine started, waiting 100ms
3	Launch second coroutine	second	50		Second coroutine started, waiting 50ms
4	Print in main coroutine	main	0	Start	Printed 'Start' immediately
5	Second coroutine delay ends	second	50	Second	Printed 'Second' after 50ms
6	First coroutine delay ends	first	100	First	Printed 'First' after 100ms
7	All child coroutines complete	main	0		Scope ends, main continues
8	Program ends	main	0		runBlocking returns, program finishes

💡 All child coroutines finished, so runBlocking scope ends and program continues.

Variable Tracker

Variable	Start	After Step 2	After Step 3	After Step 5	After Step 6	Final
Scope Active	false	true	true	true	true	false
First Coroutine State	not started	started	started	started	completed	completed
Second Coroutine State	not started	not started	started	completed	completed	completed
Output Printed			Start	Start, Second	Start, Second, First	Start, Second, First

Key Moments - 3 Insights

Why does the program wait for both child coroutines before finishing?

Because runBlocking creates a coroutine scope that waits for all launched child coroutines to complete before it ends, as shown in steps 7 and 8.

Why is "Start" printed before "Second" and "First"?

"Start" is printed immediately in the main coroutine without delay (step 4), while the child coroutines print after their delays (steps 5 and 6).

What happens if a child coroutine throws an error inside the scope?

The coroutine scope cancels all other child coroutines and the error propagates, ensuring structured concurrency and no orphan coroutines.

Visual Quiz - 3 Questions

Look at the execution table, what is printed at step 5?

A "Second"

B "First"

C "Start"

D Nothing

💡 Hint

Check the 'Output' column at step 5 in the execution_table.

At which step does the coroutine scope end?

A Step 6

B Step 4

C Step 7

D Step 8

💡 Hint

Look for the step where 'Scope ends' is mentioned in the 'State' column.

If the delay in the first coroutine was changed to 30ms, when would "First" print relative to "Second"?

A "First" and "Second" would print at the same time

B "First" would print before "Second"

C "Second" would still print before "First"

D "First" would never print

💡 Hint

Compare delays in the 'Delay (ms)' column and their effect on output timing.

Concept Snapshot

Coroutine scope creates a block where child coroutines run.
All child coroutines must finish before the scope ends.
runBlocking is a coroutine scope that blocks the thread until done.
Structured concurrency prevents orphan coroutines and manages errors.
Use launch inside a scope to start child coroutines.
Delays simulate work and show concurrency timing.

Full Transcript

This example shows how a coroutine scope manages child coroutines. The main function uses runBlocking to create a scope. Inside, two child coroutines launch with different delays. The main coroutine prints "Start" immediately. The second coroutine prints "Second" after 50 milliseconds. The first coroutine prints "First" after 100 milliseconds. The runBlocking scope waits for both child coroutines to finish before ending. This ensures structured concurrency, meaning no child coroutine is left running outside the scope. If any child coroutine fails, the scope cancels others to keep the program safe. This way, coroutines are managed cleanly and predictably.