Hi, I have to say that the original problem was NOT in precision, how I thought. The numbers are the same under NT and in ROOT/CINT. My problem was solved by replacing macro sqr(x) by function double sqr(double x). I found out in some older articles in roottalk, that macros aren't fully supported in CINT. I do apologize for wrong question. Jan.
This archive was generated by hypermail 2b29 : Tue Jan 04 2000 - 00:43:43 MET