[jberryman]

> X axis is time

That's not correct from what I'm reading.

[revelation]

If you are sampling the stack at equidistant points in time (or, sampling theorem and all, fast enough) and then determine the width on the basis of "seen in fraction of samples", that's pretty close to time. (timespan if you want)

[michael_h]

The span on the x-axis is the amount of time spent in a function, but not the passage of time left-to-right.