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

970917: netCDF 3.3.1: Integer Problem?

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 Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated.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: 970917: netCDF 3.3.1: Integer Problem?
  • Date: Wed, 17 Sep 97 10:38:00 -0600
Ralph,

> To: address@hidden
> From: "Ralph J. Hangleiter" <address@hidden>
> Subject: 3.3.1-Integer Problem?
> Organization: .
> Keywords: 199709171555.JAA27855

In the above message, you wrote:

> I'm not quite sure whether or not this is a simple question, so I decided
> against sending it to the netcdfgroup-list.
> 
>    We have been working with netcdf-2.4.3 for storing grids we use for
> continuous-fluid-dynamic calculations. No problem so far.
>    Now that we have upgraded to 3.3.1 (C-library), it seems that our DEC
> ALPHAs, running OSF4.0 (which is 64-bit), have a problem with the int's. 
> We suddenly get values which are way too big and even changing all calls
> to 3.3.1-format (checked that with the NO_NETCDF_2 define) doesn't solve 
> the problem. 
>   We know there is now 64-bit integer nctype (yet), but the 'normal'
> integer is 32bit here, so we supposed it should work.
>   Did you have any similar reports or do you have any idea what could be
> the reason for this behaviour? Or should I forward this post to the 
> netcdf-group-list?
> Thank you very much
> ralph hangleiter
> - ------------------------------------------------------------------------
>  Ralph J. Hangleiter               PGP-key        Brueder-Grimm-Allee 57 
>  address@hidden  available  D-37075 Goettingen, Germany
>  http://www.math.uni-goettingen.de/hangleit/      (+49) 551 54 16 09 

I haven't head of anything like this (and I have a DigitalUNIX/Alpha
workstation).  If the netCDF installation passed it's tests, then
it should be OK (the integer/long conversions are tested pretty
thoroughly).

It's possible that your code relies on an implicit (and incorrect)
assumption.  Can you exemplify the problem with a small piece of code?

--------
Steve Emmerson   <http://www.unidata.ucar.edu>