This study investigates the characteristics of saccadic responses to double-step target stimuli as a function of the time between the second target step and the onset of the response. The analysis suggests that goal-directed saccades are prepared in two stages: first, a decision about direction is made, which takes a variable amount of time, and second, the amplitude is calculated as an average of fixation errors. The study also demonstrates that the preparatory processes of two saccades can overlap in time, indicating parallel programming. The saccadic system continuously processes visual information, even though it responds in a discontinuous manner. A conceptual model based on internal predictive feedback and a nonlinear decision mechanism is proposed to explain the observed behavior. The study used double-step stimuli, where the second target step occurs while the subject is still preparing a response to the first step. This method has been used previously to investigate saccadic responses. The results show that the pause between saccades decreases as the interval between target steps shortens, supporting the idea of parallel programming. However, the results do not definitively prove parallel programming, as they could also be explained by grouped programming. The study found that the characteristics of double-step responses depend on the time lapse between the second target step and the onset of the response. This time lapse determines the amplitude of the first response saccade and the interval between the first and second response saccades. The study also found that the amplitude of the first saccade varies with the delay between the second target step and the onset of the response. The amplitude transition function, which describes how the amplitude of the first saccade changes with delay, was analyzed for different stimulus classes. The study also found that the interval between the two saccades of double-step responses is a function of the delay between the second target step and the onset of the response. The interval decreases as the delay increases, indicating that fewer computational steps are needed after the onset of the first saccade to generate the second saccade. The study also found that the latency and accuracy of the second saccade are influenced by the delay between the second target step and the onset of the response. The study concludes that the saccadic system can prepare two different responses simultaneously, indicating parallel programming. The results support the idea that the saccadic system continuously processes visual information, even though it responds in a discontinuous manner. The study also suggests that the saccadic system uses a nonlinear decision mechanism and internal predictive feedback to generate saccades. The findings have implications for understanding the neural mechanisms underlying saccadic eye movements.This study investigates the characteristics of saccadic responses to double-step target stimuli as a function of the time between the second target step and the onset of the response. The analysis suggests that goal-directed saccades are prepared in two stages: first, a decision about direction is made, which takes a variable amount of time, and second, the amplitude is calculated as an average of fixation errors. The study also demonstrates that the preparatory processes of two saccades can overlap in time, indicating parallel programming. The saccadic system continuously processes visual information, even though it responds in a discontinuous manner. A conceptual model based on internal predictive feedback and a nonlinear decision mechanism is proposed to explain the observed behavior. The study used double-step stimuli, where the second target step occurs while the subject is still preparing a response to the first step. This method has been used previously to investigate saccadic responses. The results show that the pause between saccades decreases as the interval between target steps shortens, supporting the idea of parallel programming. However, the results do not definitively prove parallel programming, as they could also be explained by grouped programming. The study found that the characteristics of double-step responses depend on the time lapse between the second target step and the onset of the response. This time lapse determines the amplitude of the first response saccade and the interval between the first and second response saccades. The study also found that the amplitude of the first saccade varies with the delay between the second target step and the onset of the response. The amplitude transition function, which describes how the amplitude of the first saccade changes with delay, was analyzed for different stimulus classes. The study also found that the interval between the two saccades of double-step responses is a function of the delay between the second target step and the onset of the response. The interval decreases as the delay increases, indicating that fewer computational steps are needed after the onset of the first saccade to generate the second saccade. The study also found that the latency and accuracy of the second saccade are influenced by the delay between the second target step and the onset of the response. The study concludes that the saccadic system can prepare two different responses simultaneously, indicating parallel programming. The results support the idea that the saccadic system continuously processes visual information, even though it responds in a discontinuous manner. The study also suggests that the saccadic system uses a nonlinear decision mechanism and internal predictive feedback to generate saccades. The findings have implications for understanding the neural mechanisms underlying saccadic eye movements.