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

Re: netCDF-3 ncdump bug

Hi Mark,

> >> On Thu, 22 May 1997 13:16:38 -0600,
> >> Russ Rew(RR) wrote:
> 
> RR> The next minor release will probably come out next week.  It will
> RR> have bug fixes to the ncgen "-f" and "-c" options, some ports to a
> RR> few additional platforms, miscellaneous fixes, and a new
> RR> RELEASE_NOTES document.
> 
> Just a status check. Some of our users are "antsy" to get the improved 
> performance benefits of the latest version. Is the next minor release
> right around the corner? Or did something else come up to push the
> release back?

That's spelled "ANSI", and it's the kind of compiler you'll need for
version 3.3.1 :-).

After all these years, you'd think I would learn that unanticipated
problems always come up to delay these releases beyond my wildest
over-estimates.  Since I wrote the above, we've had to rewrite much of
the Fortran interface implementation and ncgen's -f option to make
things work on various Cray systems.  We've also just finished testing
things on Win32 platforms, with all necessary changes rolled in.  I've
appended the current version of the release notes.  As far as I know,
it's just a matter of packaging and one last round of testing at this
point.  There's a good chance we'll have it ready in two or three days.

--Russ

_____________________________________________________________________

Russ Rew                                         UCAR Unidata Program
address@hidden                     http://www.unidata.ucar.edu

This file contains a high-level description of this package's evolution.
Entries are in reverse chronological order (most recent first).


VERSION         COMMENTS
-------         --------

3.3.1   YYYY-MM-DD

                The FORTRAN interface should now work on more systems.
                In particular:

                    It should now work with FORTRAN compilers whose
                    "integer*1" datatype is either a C "signed char",
                    "short", or "int" and whose "integer*2" datatype is
                    either a C "short" or "int".

                    It should now work with FORTRAN compilers that are
                    extremely picky about source code formatting -- in
                    particular, it should now work with the NAG f90
                    FORTRAN compiler.

                The dependency on the non-POSIX utility m4(1) for
                generating the C and FORTRAN manual pages has been
                eliminated.

                EXTERNAL statements have been added to the FORTRAN
                include-file "netcdf.inc" to eliminate excessive
                warnings about "unused" variables (which were actually
                functions) by some compilers (e.g. SunOS 4.1.3's f77(1)
                version 1.x).

                Building the netCDF-3 package no longer requires the
                existence of the Standard C macro RAND_MAX.

                Fixed an ncdump bug resulting in ncdump reporting
                    Attempt to convert between text & numbers
                when _FillValue attribute of a character variable set to
                the empty string "".

                Made ncgen tests more stringent and fixed various bugs
                this uncovered.  These included bugs in handling byte
                attributes on platforms on which char is unsigned,
                initializing scalar character variables in generated C
                code under "-c" option, interspersing DATA statements
                with declaration statements in generated Fortran code
                under "-f" option, handling empty string as a value
                correctly in generated C and Fortran, and handling
                escape characters in strings.  The Fortran output under
                the "-f" option was also made less obscure and more
                portable, using automatic conversion with netCDF-3
                interfaces instead of "BYTE", "INTEGER*1", or
                "INTEGER*2" declarations.

                Made ncgen "make test" report failure and stop if test
                resulted in a failure of generated C or Fortran code.