18 Apr
2007
18 Apr
'07
9:41 p.m.
The trouble with this (exponentially increasing) de Bruijn construction is that --- as might be expected --- it's insufficiently constrained, which becomes immediately apparent on inspecting the length 65536 segment. For instance, using the numerically earliest possible span-16 deBruijn sequence, the sum of the first 1024 digits is only 228. For fixed word size, such imbalances will only worsen as as the construction proceeds. Galois sequences have to be a better prospect from this point of view. WFL