> Date: Wed, 9 Jun 1999 17:39:53 PDT > From: Chris Peters <[log in to unmask]> > I suppose mathematically you could write base-pi numbers > using only the digits 0, 1, 2 and 3, but what kind of crazy stuff > would it give us? Stuff like 2 + 2 = 10.2201220211211103010+... Doing carries when the partial result is 301102111002022211300 or larger, or living with ambiguous representations... Headaches. > I do see its mathematical application ... maybe a > geometrically-inclined conculture would use it? There aren't that many powers of pi in geometry. Not until you get to higher-dimensional geometry anyway. What I could imagine is a language where numbers are easily converted one factor of pi up or down by tacking on a particle: 'two diam foot' is just over 6 foot. E.g., used to express the length of string you need to hold a two foot thick bundle of something. 'seven circ yards' is about the same length. But used for instance when you have measured a tree with a tape around it, and need to express how thick it is. The point is that people would get used to estimating sizes in all three scales ('diam', plain and 'circ'); perhaps things like tape measures would be marked that way too. So no need to multiply or divide in your head in these cases, just think of the same number in the other scale. Lars Mathiesen (U of Copenhagen CS Dep) <[log in to unmask]> (Humour NOT marked)