The usual reason given is that solar radiation would otherwise be able to penetrate the hull of the vessel, and cause us all to die... and if our windows were several feet thick, they would shatter every time a micrometeorite hit them.

Come now, to shield against radiation, you may need a few meters of material, but it doesn't have to be glass. Indeed, water does a much better job.

However, the risk of having windows of any kind pitted and broken by micrometeorites is a very real concern. How to get light in then?

Well, why not pipe it in via fiber optics. Natural sunlight is already filtered by several miles of atmosphere, so if the fibers are not covering 100% of the colony, that should be no major problem. Fibers a few feet thick are no problem to make, indeed, there may be a few suitable leftovers from cables lying around already, and as for shielding, well, we should be able to make fiberoptics out of tinted glass just as easily as we make tinted windows out of tinted glass, right?

Fiber optics have a number of distinct advantages over windows. In areas distant from the sun, we can attach large lenses to the ends of the cables, and collect more sunlight. In areas near to the sun, we can have fewer cables, and collect less sunlight. The thin diameter of a fiber makes for a smaller hole if one is "broken," and it can be pushed through several airtight layers, wheras a window would need to replace some of the shielding.

The cables can be bent about, to allow micrometeorite shielding to overlap all areas of the colony, whereas a window would need to open right up to exposed sunlight. In fact, for some applications, layers of cloth are used for micrometeorite shielding, and fiberoptic cables might be woven into cloth to provide this shielding as well as giving light.