Good question. The pattern will indeed turn sideways. 1. There is nothing deep about the shape of the slit. You can do the "double-slit" experiment with "doubl

Good question. The pattern will indeed turn sideways. 1. There is nothing deep about the shape of the slit. You can do the "double-slit" experiment with "double-pinhole" or "double-weirdshapes" and you will get the same behavior, just that the math will be different. 2. In fact, there is a simpler experiment you can do which gives the same qualitative results, but in which the shape of the wave is much less relevant. You can split a single photon into two fiber cables using a beamsplitter, then interfere the light in those cables using another beamsplitter. The good thing about this is that all the operations in this setup are linear, because light is not lost anywhere. In a double-slit experiment, light that falls outside the slits gets blocked and so the operations are non-linear and you have to worry about the shape of the slits and wave a lot more. 3. I am not quite sure which two dimensions you mean, when you say 2d wave. There are two axes in the plane of the slits (longitudinal) and another axis perpendicular to this plane (transverse). The wave of light is in fact 3D, but lets simplify. If its 2D in the longitudinal axes, then obviously the shape/orientation of the slits matter. Imagine putting a rainbow colored piece of paper against the slits. Depending on the orientation of the slits, different colors will make contact with the slits. If its 2D in transverse+one longitudinal axis then it still matters. Imagine taking a stack of identical rainbow colored paper and put it edge on against the slits. If you rotate the slits, obviously the colors making contact with the slits will change. Hence, the interference pattern changes again.