It isn't a random pattern, but yes. A logarithmic expansion of baseline distances seems to be the first thing you want in a sparse array. Ideally you want to b

It isn't a random pattern, but yes. A logarithmic expansion of baseline distances seems to be the first thing you want in a sparse array. Ideally you want to be able to look in all directions with about the same amount of sensitivity and angular resolution. But the UV coverage of even an asymmetrical, relatively sparse array like the VLBA is surprisingly good, because as the Earth rotates, the antennas describe sweeping arcs on the sky. The regular UV coverage of the VLA is great, though, for sensitivity at short duration observations. But Arecibo was great for pulsar work. :-(