> It should strip all trailing zeros from the result:

Hmm.  Thinking about this some more, I don't think this is true:  format() shouldn't be throwing away significant information (in Decimal 
the number of trailing zeros *is* significant information) unless that's 
necessary to fit the result into the given precision.

So I think the current code is correct.

However, we could do with documenting the precise rules being used for 
Decimal formatting somewhere.