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

960118: NETCDF fortran errors with linux

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 Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated.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: 960118: NETCDF fortran errors with linux
  • Date: Thu, 08 Feb 96 08:18:15 -0700
Darin,

>Date: Wed, 07 Feb 1996 21:13:13 -0500 
>From: Darin Ernst <address@hidden>
>Organization: PPPL
>Subject: Re: 960118: NETCDF fortran errors with linux 
>Keywords: 199602080214.AA03091 

In the above message you wrote:

> I have been using netCDF successfully for a couple of weeks. I use
> it to create a file of about 30 large arrays.  This works well,
> at least after compressing with gzip (875 vms blocks down to
> 60 blocks ~30kB).  Many of the arrays are empty.
> 
> Here are some observations:
> 
> 1) The byte type does not seem to work on fortran type logical.

My understanding is that the FORTRAN-77 data types LOGICAL, REAL, and
INTEGER must be the same size in order to make EQUIVALANCE work
correctly.  (This does not apply to types like LOGICAL*1).  If you want
a single-character data type, then I suggest using LOGICAL*1.

> 2) I could not compile netCDF on a VMS alpha AXP machine, even
>     when I used the /standard=vaxc compiler flag.  The compile
>     on a vax went fine.

We can't test the netCDF package on a VMS Alpha because we don't have
access to one.  In this situation, we depend on others figuring out what
needs to be done to get the package to build and work.

> 3) I would find it very useful to have the default limit on
>    the number of arrays to be increased beyond thirty or so.
>    We have an application with about 2000 arrays of 20 points each.  
>    Currently I get around this by equivalencing these arrays to
>    a single large array.  I believe I had a problem reading more 
>    than 30 arrays even though this limit is not 'enforced'.

To get the netCDF package to support more than the default number of
variables, it is only necessary to edit the file libsrc/netcdf.h and
rebuild the package.  Change the value of the macro MAX_NC_VARS to whatever
you desire.

I believe the next release of the netCDF package will support 2000
variables by default.

> Have you had any success on VMS alpha machines?

We don't have access to one.

> Thanks,
> 
> Darin Ernst
> address@hidden

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