I forwarded this thread to John Barnes, so it ought to get fixed when he 
gets around to revising the existing part of the Rationale. FYI, I sent him 
a list of 51 bugs I found when I created (the not yet finished) HTML 
version, so it's not quite a foolproof document at this time. :-)

                                    Randy.

"Nasser M. Abbasi" <nma@12000.org> wrote in message 
news:jrsl7g$33m$1@speranza.aioe.org...
> On 6/20/2012 5:44 AM, Egil H�vik wrote:
>> On Wednesday, June 20, 2012 12:09:00 PM UTC+2, Nasser M. Abbasi wrote:
>>>    function T(N: Natural) return Long_Long_Float is
>>>     (if N = 0 then 0.25 else
>>>        (T(N�1)�2.0**(N�1)*A(N�1)�A(N))**2);
>>
>> Should be:
>>
>>     function T(N: Natural) return Long_Long_Float is
>>      (if N = 0 then 0.25 else
>>         T(N�1)�2.0**(N�1)*(A(N�1)�A(N))**2);
>>
>
> thank you. I verified that I get Pi now with the above change.
>
> May be someone at Adacore can update the pdf file on that web site,
> as it is not good to have this typo in the code given it is an
> Ada 2012 document.
>
> --Nasser
>