User blog comment:LittlePeng9/Random Turing machines/@comment-5150073-20130427202925/@comment-1605058-20130428085920

When programming machine I assumed that for every {a,b,c,#} array it is always true that {a,b-1,c,#}>b. For a=1 it isn't true, and I missed it. I also fixed machine for arity 1.

Now only problem I see is that output of a^1 is always decreased by 1. But this is minor problem, as it only appears for {a,1} arrays.