[ yet another pi calculation program in bf

  Just like for pi16.b the accuracy of the result depends on the cellsize:

   - using  8 bit cells causes an overflow after 4 digits
   - using 16 bit cells causes an overflow after 537 digits
   - using 32 bit cells causes an overflow after several millions of digits

  It's about ~38 times shorter than pi16.b, ~364 times faster and works with
  not-wrapping (bignum) implementations.

  by Felix Nawothnig (felix.nawothnig@t-online.de) ]

>  +++++ +++++ +++++ (15 digits)

[<+>>>>>>>>++++++++++<<<<<<<-]>+++++[<+++++++++>-]+>>>>>>+[<<+++[>>[-<]<[>]<-]>>
[>+>]<[<]>]>[[->>>>+<<<<]>>>+++>-]<[<<<<]<<<<<<<<+[->>>>>>>>>>>>[<+[->>>>+<<<<]>
>>>>]<<<<[>>>>>[<<<<+>>>>-]<<<<<-[<<++++++++++>>-]>>>[<<[<+<<+>>>-]<[>+<-]<++<<+
>>>>>>-]<<[-]<<-<[->>+<-[>>>]>[[<+>-]>+>>]<<<<<]>[-]>+<<<-[>>+<<-]<]<<<<+>>>>>>>
>[-]>[<<<+>>>-]<<++++++++++<[->>+<-[>>>]>[[<+>-]>+>>]<<<<<]>[-]>+>[<<+<+>>>-]<<<
<+<+>>[-[-[-[-[-[-[-[-[-<->[-<+<->>]]]]]]]]]]<[+++++[<<<++++++++<++++++++>>>>-]<
<<<+<->>>>[>+<<<+++++++++<->>>-]<<<<<[>>+<<-]+<[->-<]>[>>.<<<<[+.[-]]>>-]>[>>.<<
-]>[-]>[-]>>>[>>[<<<<<<<<+>>>>>>>>-]<<-]]>>[-]<<<[-]<<<<<<<<]++++++++++.