Recently physicists have managed to build "attosecond lasers", lasers which emit pulses 10^-18 seconds long, interrupted by much longer periods of darkness (at least 10^-14 seconds). Before them, lasers emitting femtosecond (10^-15 seconds) pulses have been around. Assuming they produce visible light, what colour is it?

I was thinking about this a little more and I think that the answer depends on the duration of the periods of darkness between pulses.

This is because the the periods of darkness determine the fundamental frequency f0 which is used to determine the components in the Fourier series.  The "discreteness" of the spectrum depends on this.

For example, if the period between pulses is 1 milisecond then the spacing of the discrete frequencies is only 1000Hz.  Compared to 10^14Hz, 1000Hz spacing is virtually 0.  So, the spectrum looks "white" because there is such a small spacing between the frequency components.

However, if the periods of darkness are 10^-14s then the spacing of the frequencies is on the order of 10^14Hz.  So, this leaves a small number of discrete frequencies in the visible.  This is why I was able to determine a "color".

Consequently, I think if the periods of darkness are something much longer than 10^-14s (like 10^-10s) then there are many frequency components which are in the visible and the spectrum looks more "white" as in the solution.