Press the stop sign. Notice how the number of times the project has reported itself stopped never goes above 1, because it doesn't ever realize it was stopped a second (or third, or fourth, etc.) time, because it was truly stopped. You can also see this in that the variable named "some variable" stops incrementing the second time you click the stop sign, but not the first (because the code incrementing it gets restarted after the first time, but not the second).
Traditional stop sign clicked blocks work by resetting the timer in a forever loop and using the when timer > 0 hat block to restart the next frame. This works by having an absurdly long time to wait for the timer, enough to basically guarantee that anyone who clicked the stop sign will have either given up or died before that happens. The variable this is stored in is only set to this value when the stop sign is first clicked, which is then added to 0 (the time the project normally waits after the stop sign is clicked to restart itself) in the hat block detecting when the timer reaches above a certain amount. As a result, every subsequent time the stop sign is clicked, no code actually runs and the project is effectively stopped. If you're wondering what this is useful for, the answer is the very few cases where you need some code to run before the project is stopped completely (such as prompting the user to copy a save code so that they don't loose their progress), but don't want to run the project endlessly.
