[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
20001220: HPUX trouble with sfchck, dcwatch dcshef
- Subject: 20001220: HPUX trouble with sfchck, dcwatch dcshef
- Date: Fri, 22 Dec 2000 13:37:29 -0700
On Wed, 20 Dec 2000 address@hidden wrote:
Thanks very much for this.
I still have problems compiling dcshef under comet and dcwatch under
ldmConnect, but I do not need either of them, so I bypass them.
The sfchck still gives the same segmentation violation - maybe once I
have upgraded to HPUX-11.00 everything will work out. I'll be doing that
first thing in January and will let you know how it works out.
The rest works just great and is appreciated!
Thanks again for everything and Season's Greetings.
Stephen Sinnis | Tell:905.566.9511x379
Pelmorex Inc.
*****************************************************************
Stephen,
In sfchck.f, there is a large array defined:
LOGICAL datflg ( LLSTFL, LLMXTM )
with my defaults of LLSTFL = 29700 and LLMXTM = 300, this
probably causes the segmentation fault on startup, before anything
in executed. You can try replacing instances of LLSTFL in this
program with a smaller number of say 5000 if you don't have more than
that number of stations in your surface files. This will let
you use this program. Otherwise, you can compile the entire
source of GEMPAK using a smaller LLSTFL in the GEMPRM file.
In $NAWIPS/unidata/ldmbridge/dcwatch, the Makefile needs $(SYSLIBS)
when linking (this will resolve the math functions from -lm).
eg:
$(PROG) : $(PROG_OBJS)
$(LINK.f) $(NOFORMAIN) $(PROG_OBJS) $(LIBS) $(SYSLIBS) -o $@
(I found that when make failed in dcwatch, it also skipped building in
$NAWIPS/unidata/programs).
I haven't found anything in compiling dcshef as of yet.
Steve Chiswell
Unidata User SUpport