Convolution followup

Hi Fred,

Thanks a lot for the prompt response. I've now got the hang of it.
It is reassuring that my thought process is compatible with your
clarification, save for the incorrect semantic usage of "present" to 
state the input x(n) and response h(n). To re-state what you've said:
x(1) has to butt(align) with h(1). 

Another question: Why use dummy variable Tau? How does one arrive at 
h(t-tau) and not h(tau-t)on the tau-axis?

Regards

Azhar