Cees de Groot wrote:
> I *understand* it, but it's not what I *want* :).

The point is that, there is no reason why 0 bitInvert should return 1 
rather than 3, 7, or 15.

Your data type is a BitArray, but if you encode it as an Integer, that's 
not enough.  You need another instance variable to store the width or 
alternatively a mask (i.e. (1 bitShift: width) - 1).  Then, you have 
something like this:

(the bitOr:'s in the masks are not typos).

& aBitString
     ^BitString
         value: (self value bitAnd: aBitString value)
         mask: (self mask bitOr: aBitString mask)

| aBitString
     ^BitString
         value: (self value bitOr: aBitString value)
         mask: (self mask bitOr: aBitString mask)

invert
     ^BitString
         value: (self value bitXor: self mask)
         mask: self mask

xor: aBitString
     ^BitString
         value: (self value bitXor: aBitString value)
         mask: (self mask bitOr: aBitString mask)

xnor: aBitString
     ^(self xor: aBitString) invert

count
     | n cnt |
     n := self value. cnt := 0.
     [ n = 0 ] whileFalse: [
         cnt := cnt + 1.
         n := n bitAnd: n - 1 ].
     ^cnt

Paolo