It seems to be foolish to use an even number, such as n=8, in the code of “projection method for 5-fold …“. Note, that I built the code for odd n, especially n=5. For n=8 we get only four different directions and pairs of sets of equidistant parallel lines having the same orientation. Thus we simply get four different sets of parallel lines with two different spacings, which alternate. However, I found a good value for the offset s, resulting in an interesting quasiperiodic design:
It faintly resembles the design I got earlier with iteration, see “Another tiling with octagons“. Is it related to the Ammann-Beenker tiling ? Can you get nice results for n=12 too?