comppi



I am interested in the underlying theorem for the computation of pi in comppi in float.lisp. 
Perhaps I am too stupid to recognize it or simply I don't know it.
Can anybody help?
Thanks in advance

Volker van Nek