> However I do not think the idea to keep the Fortran environment just to be able to call That's right. It would be better to implement numerical methods in C++, but in the current situation it is much more cost effective to use existing code, even if it comes from fortran. > However NTFS serves me well. Except that you cannot use my example :-) > One can find this information within well-known "cfortran" packages. > (...) > cfortran.h gen.h higz.h kernlib.h minuit.h Cfortran is quite nice, but ... after I looked into rkstp/deqbs/deqmr definitions in gen.h and kernlib.h I immediately decided to define them myself (just have a look into these files and you'll understand :-) . The approach I used works in most cases, except in some very rare ones, like Windose, which I anyhow do not intend to use (nor anyone I know :-) . In any case, if there are many people interested in providing an interface from CERNLIB to root, one could discuss/implement a more general solution (either by improving these CERNLIB "*.h" files, or by managing interfaces by hand like, for example, in the h2root). Jacek.
This archive was generated by hypermail 2b29 : Tue Jan 02 2001 - 11:50:39 MET