[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

970120: Problem report: netcdf-3.3a

This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.

  • Subject: 970120: Problem report: netcdf-3.3a
  • Date: Tue, 21 Jan 97 09:56:39 -0700
Russ,

Harvey wrote:

> > 2. The file INSTALL should mention:
> > 
> >    (a) How to build & test system without FORTRAN. (This could
> >        be made a little simpler if all you had to do was specify
> >        FC=''. Also it would be nice if this happened automatically
> >        on system which did not have any FORTRAN.) Note that test of
> >        ncgen includes FORTRAN test for -f option.

Odd.  As far as I can tell from looking at the top-level Makefile and
ncgen/Makefile in the distribution, support for ignoring the FORTRAN
interface (via an empty FC variable) is already there.

Harvey, doesn't it work?

> >    (b) Default for -prefix is ../.. (relative to subdirectory e.g. libsrc
> >        i.e. .. relative to src)

I'm afraid I don't see the problem.  The default installation prefix
was *designed* to be the absolute pathname equivalent to `..' relative
to the top-level source directory.  This is the standard way to
install a "package".  The installation prefix can be overridden via
the `-prefix=<path>' option of the configure script.  I checked the
distributed configure script for this feature and it worked.

--------
Steve Emmerson   <address@hidden>