Performing spectral analysis on a signal with the FFT (Fast Fourier Transform) requires the signal to be cut up into "chunks." The process of extracting one of those chunks is often called "windowing." It's as if you're looking at a portion of the signal through a window, and ignoring the rest of the signal. Sometimes, it's important to shape the portion of the data that lies inside the window, because it can favorably affect the result of the FFT.
The "Flat Top" window type, for example, works really well for accurately measuring the amplitude of a signal. The "Rectangular" or "Uniform" window type does not shape the data within the window at all. It offers the finest frequency resolution in the spectrum, but it may not accurately preserve the signal's amplitude at certain frequencies. There are other trade-offs associated with the various window types.
A chart of frequency-domain effects of several different window types can be found
here. Notice that the Rectangular and Flat Top windows lie at the extremes, in terms of frequency resolution, with everything else lying somewhere in between them.
