[GRASSLIST:3260] RE: NVIZ2.2 compile error TkIntDecls.h redux

[John Gillette]

It looks like a tcl problem to me.  (Just guessing, I am not a
programmer).  The scripts subdirectory is full of the tcl script
files.

The "fix" I used of renaming the header files worked for me BUT
I don't think this is a good fix because the headers COULD be
different for different versions of tcl/tk.  See [GRASSLIST:3100]
in which Markus says "o.k., I have added the missing files. 
Now we have:" and goes on to list the changes made. I assume he 
ment that he added these into the CVS file.  Check the CVS 
RELEASE branch snapshot (NOT experimental branch).  All you need 
is the correct header file (I hope). Try un-taring the snapshot
and search for the headers in .../src.contrib/GMSL/NVIZ2.2/src/

If this doesn't fix it, you will need more "expert" help than
mine.  If you need more help include information on OS and
library versions. 

John Gillette

> -----Original Message-----
> From: Andrew [mailto:andrew at quakes.potsdam.edu]
> Sent: Thursday, February 28, 2002 1:37 PM
> To: John Gillette
> Subject: Re: [GRASSLIST:3258] RE: NVIZ2.2 compile error TkIntDecls.h
> redux
> 
> 
> 
> > I have these suggestions:
> > (1) look at your region settings.  Too large a region will slow
> >     NVIZ down considerably.  I made this mistake the first time I
> >     ran it.  Try something small like 501 X 501 rows and cols
> 
> 675 X 500, .01 degrees separation in both x and y for 
> coverage of NY state
> and eastern Pennsylvania.
> 
> 
> > (2) Try a CVS snapshot to compile.  I believe Markus fixed the
> >     header file issue for the compile issue.
> 
> I fixed the include file problem before I wrote. It was an 
> obvious "rename"
> based upon the information in the files listed in the error 
> message. I found
> Marcus' response on the list when I searched on the problem 
> and recompiled
> one more time. I have never worked from CVS, I try to always 
> take the latest
> stable package of things and then search the recent lists if things go
> phlooey. I have found no mention of this particular behavior 
> to suggest CVS
> should be necessary. I don't mind learning new things, but I am bit
> overextended ...
> 
> 
> 
> 
> > (3) Tell us exactly what messages you get as it starts up.  I
> >     get the following:
> > Loading Data
> > Loading Data
> > translating colors
> > Adding panels from /scratch/grass5p3/grass5/etc/nviz2.2/scripts
> 
> Exactly this much. The program is still active, appears as a 
> process, can be
> paused and continued indefinitely, can be exited by right 
> clicking the start
> up form, etc,  but nothing else happens and nothing dies on 
> its own as noted
> previously. I tried to start from the grass command line and 
> got no panels,
> but the program proceded to the same place before "losing 
> contact" with my
> need.
> 
> There were no other errors or significant warnings that I 
> noticed during the
> recompiles.
> 
> >
> > I have attempted to start nviz from the TclTkgrass 
> interface (and from
> > the grass command line). The program starts and opens the "waiting"
> > window, prints the line about loading the panels from the 
> index (default
> > /usr/local/grass5/etc/nviz2.2/scripts/panelIndex) and goes into an
> > infinite do-nothing mode. I can stop and continue from the startup
> > interface button, exit by right clicking, but nothing else happens.
> >
> > Suggestions from anyone?
> >
>