> I have committed (revision 9319) a general fix for this issue which
> involves
> "call flush(6)". I have checked in the appropriate documentation that the
> flush routine is available in fortran support libraries for the g77,
> gfortran, absoft, intel, and Portland group compilers so I think this fix
> should probably be okay on most if not all platforms. However, I don't
> know
> whether "call flush(6)" works for the native Windows fortran compiler so I
> hope those with access to that compiler will comment.
>
Hi Alan,
I have tested if "call flush(6)" is accepted by the Compaq Visual Fortran
compiler (version 6.6C) and the Intel Fortran compiler on Windows.
While my little test program presents the output in the same way with
and without the "call flush(6)" statement, at least it compiles and links
with both compilers.
There may be some compilers out there for which this does not work -
I know of the SGI Fortran compiler that uses (used?) unit number 101
instead of 6 for this flushing operation. But that is a mere anomaly
that I would not worry about.
(The issue of buffering occurs also in the rxvt console that comes
with an installation of MSYS on Windows - that makes that console
rather hard to use with interactive programs.)
Regards,
Arjen
----
Test program:
! tryflush.f90
! Does flush work on stdout for different compilers?
!
program tryflush
write(*,'(a)', advance = 'no') 'Type in a number:'
read(*,*) number
write(*,*) 'Answer: ',number
end program
------------------------------------------------------------------------------
This SF.net email is sponsored by:
SourcForge Community
SourceForge wants to tell your story.
http://p.sf.net/sfu/sf-spreadtheword
_______________________________________________
Plplot-devel mailing list
Plplot-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/plplot-devel
