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

20000321: gempak5.4 installation



Mike,

Your procedure to build in a gempak3 parallal directory is fine.
When complete, you can move your current /usr/local/gempak to
something like /usr/local/gempak.old and then move your
/usr/local/gempak3 to /usr/local/gempak. Quick and easy.
All you have to do is redefine the NAWIPS variable in the Gemenviron
file when you move the directories around.

The ACARS and PROFILER decoders need the NetCDF library if you
want those data sets (dcacars and dcncprof). All you have to do 
is define the location of the NetCDF header and includes NETCDF and NETCDFINC in
the appropriate location in $NAWIPS/config/Makeinc.common.

I added them to the default build since most sites were now getting at least 
the FSL2 feed.  If you don't want to build these decoders, then edit the 
$NAWIPS/unidata/ldmbridge/Makefile and take them out of the DECODERS list
at the top. 

The WANTCDF variable only deals with the ability to read NetCDF files
in the gd programs. Nothing to worry about here defined as NO.

The "blank" definitions are fine the way they are. Some of these get
overridden by the Makeinc.platform file that is included at the end
of the Makeinc.common file. Since not all platforms need extra
compiler and path definitions, these will have empty declarations on some 
platforms.

Steve Chiswell
Unidata User Support


>From: Michael Voss <address@hidden>
>Organization: .
>Keywords: 200003212015.NAA28268

>Hello (Steve),
>
>
>I have an older version of Gempak-5.4 running on solaris 2.6. I am
>finally getting around to upgrading with the P116 distribution. I have
>the old gempak in /usr/local/gempak. I see that the thing to do from the
>archives and web documentation is to start with a clean top level. Well,
>being the cautious sort I decided to make a clean install from another
>top level /usr/local/gempak3 (crazy I know, but bare with me). My
>thinking is that I can get some of my trouble-shooting out of the way
>without interupting the current data ingestion and processing. If this
>works then I can go ahead an mv the old /usr/local/gempak to
>/usr/local/gempak.old and then install the new p116 distribution in
>/usr/local/gempak, knowing that it should work :) 
>
>
>Anyway, I tried to install in a clean /usr/local/gempak3 and it seems to
>do most of the make until I got the following error related to netcdf
>include file:
>
>
><bigger>f77 -O  -c dcsels_out.f
>
>dcsels_out.f:
>
>        dcsels_out:
>
>"gemprm.prm", line 20: Warning: local variable "ermiss" never used
>
>cc -DUNDERSCORE -DSunOS -O  -c -DGEMPAK -DSTDC -D_DEV_CONSLOG
>read_wwus60.c
>
>-n making in
>
>/usr/local/gempak3/unidata/ldmbridge/dcncprof
>
>cc -DUNDERSCORE -DSunOS -O  -c -I/upc/netcdf/include -DGEMPAK -DSTDC
>-D_DEV_CONSLOG dcncprof.c
>
>cc -DUNDERSCORE -DSunOS -O  -c -I/upc/netcdf/include -DGEMPAK -DSTDC
>-D_DEV_CONSLOG decode_ncprof.c
>
>"decode_ncprof.c", line 3: cannot find include file: <<netcdf.h>
>
>"decode_ncprof.c", line 27: undefined symbol: NC_GLOBAL
>
>cc: acomp failed for decode_ncprof.c
>
>*** Error code 2
>
>make: Fatal error: Command failed for target `decode_ncprof.o'
>
>Current working directory /usr/local/gempak3/unidata/ldmbridge/dcncprof
>
>*** Error code 1
>
>make: Fatal error: Command failed for target `all'
>
>Current working directory /usr/local/gempak3/unidata/ldmbridge
>
>*** Error code 1
>
>make: Fatal error: Command failed for target `all'
>
>Current working directory /usr/local/gempak3/unidata
>
>*** Error code 1
>
>make: Fatal error: Command failed for target `all'
>
>vortex.sjsu.edu%
>
></bigger>
>
>questions:
>
>I thought that if the WANTCDF definition is set to NO in the
>Makeinc.common file then I don't need to worry about netcdf, or does this
>indicate some other problem?
>
>
>Also, I noticed in the Makeinc.common file that the were two entries
>without defaults:
>
>
>LDMINC          =
>
>LDMLIB          =
>
>BRIDGE          = $(LIBDIR)/bridge.a
>
>MANDIR          = $(NAWIPS)/man
>
>UNILIB          = $(NAWIPS)/unidata/lib/$(NA_OS)
>
>
>does this matter for anything?
>
>
>thanks for your help,
>
>
>Mike
>
>--------------------------
>
>Mike Voss                                
>
>Department of Meteorology              
>
>San Jose State University                        
>
>One Washington Square                             
>
>San Jose, CA 95192-0104   
>
>             
>
>408.924.5204 voice
>
>408.924.5191 fax   
>