From: Alan W. I. <ir...@be...> - 2006-05-16 19:43:00
|
Arjen, I need a quick response to the question at the end... Alan On 2006-05-16 09:03+0100 Andrew Ross wrote: > So if we can sort out this .f95 / .f90 issue then the F95 bindings > should work for at least 2 of the main commercial fortan compilers, > provided of course that F77 and FC are both set to the same thing. That's wonderful news, Andrew. I have reviewed the various pieces of documentation for the Intel, Absoft, Portland Group, and gfortran compilers. They all claim to be fortran 95 compilers. Also, Absoft, Portland Group (at least its current release), and gfortran accept the .f, .f90, and .f95 suffix. (Fortran 95 code can appear with all suffixes. The .f suffix simply means the code is in fixed format and the .f90 or .f95 suffixes mean it is in free format.) Despite being fortran 95 capable, the Intel compiler only accepts .f and .f90 suffixes (unless you use the impossible [from the autotools point of view] -TfFILENAME option to tell the compiler to not worry about suffixes for the file called FILENAME.) So it appears the the current Intel compiler (unless you use the impossible -Tf option) and the old release of the Portland Group compiler do not accept the .f95 extension, but everything else I have looked at does. Also, it appears that all these fortran 95 compilers accept the .f90 extension. (I have just checked that with gfortran in bindings/f95, and Andrew reports good results for the Intel compiler and old release of the Portland Group compiler.) To allow our users to use the Intel compiler and the old Portland Group compiler, I am leaning toward leaving our f95 directory names intact (to indicate we really do have a fortran 95 interface), but renaming all the present *.f95 files in bindings/f95 and examples/f95 to *.f90 as suggested by Andrew. Arjen, you have a lot more fortran 95 experience than me. Also, you are the original instigator of the fortran 95 interface. Thus, I think it is your call. What do you think is the best thing to do? Alan __________________________ Alan W. Irwin Astronomical research affiliation with Department of Physics and Astronomy, University of Victoria (astrowww.phys.uvic.ca). Programming affiliations with the FreeEOS equation-of-state implementation for stellar interiors (freeeos.sf.net); PLplot scientific plotting software package (plplot.org); the Yorick front-end to PLplot (yplot.sf.net); the Loads of Linux Links project (loll.sf.net); and the Linux Brochure Project (lbproject.sf.net). __________________________ Linux-powered Science __________________________ |