Looking back on it, I suppose the failure of the Pangram Machine Mark I was piece of good fortune. I mean, otherwise the faster Mark II model might never have come into existence. Another advantage of the latter was that different initial text constraints of rewiring. This meant I could experiment at will, confined only by the initially chosen set of number-word ranges, an important limitation, but still granting much scope.
"In Quest of a Pangram (Part 2),"
Word Ways: Vol. 25
, Article 5.
Available at: http://digitalcommons.butler.edu/wordways/vol25/iss2/5