This is all really hard stuff, and speculation. Normal numbers are irrational numbers that have the property that all the digits in their decimal expression are equally distributed. http://mathworld.wolfram.com/NormalNumber.html An example of a normal 10-number might be: 0.123456789101112131415161718192021222324252627... Clearly whether a number is normal can depend on the base that it is represented in, so it makes sense to refer to b-normal numbers where b is the base. An absolutely normal number is a number that is normal in any fixed base. I was wondering about this stuff a while ago, and the following questions came to me - insightful comments would be appreciated: If a number is [itex]p[/itex]-normal for all primes, is it necessarily absolutely normal? If a number is [itex]n^i[/itex]-normal for some whole number [itex]n[/itex] is it absolutely normal? If a number is [itex]p[/itex]-normal and [itex]q[/itex]-normal is it also [itex]pq[/itex]-normal? What about the converse? What about more exotic bases - like negative bases or base factorial? I'd like to call a number !-normal if its base factorial expansion has a distribution so that all digits have zero density, in all of it's base factorial expansions. (Phrased that way to cover 0.1,2,3,4,5,6,7,8,9,10,11,12,13...=1.0,0,0... (base factorial) and it's cousins.) Base factorial seems particularly interesting because it should dovetail well with the Taylor-series-type stuff we have for caculating constants.