grass5 and NULL was:Re: NVIZ under Solaris?

[Markus Neteler]

Hi Malcolm, dear GRASS community,

On Wed, Oct 18, 2000 at 10:35:18AM -0500, Malcolm Williamson wrote:
> Hi,
> I'm posting to the list in case other folks share the same question. I'm
> using the Solaris binary release for GRASS 5.0 beta 8, and would like to
> incorporate NVIZ. Although the man page and the tcltkgrass module are
> there, the program itself does not appear to be. Is there a compiled
> version of NVIZ for Solaris floating around, or will I have to compile
> myself?
Yesterday we could fix a bug when running NVIZ with tcl/tk8.3 libraries.
Due to the updated "configure" routine it should be possible to compile
the current GRASS5/CVS version on SUN including the NVIZ.
As I am just compiling it on SUN, I may be able to provide NVIZ binaries
the next days.

> Also, g.manual appears to be broken in this particular binary release (I
> ended up stealing the appropriate executables from beta 7). I get the
> following:
> GRASS 5.0beta8 > g.manual g.list
> sh: /proj/grass4.1/grass5.0/grass5.0beta_sol26/man/1/g.list: cannot
> execute

This should be fixed already. For you: To fix it easily, edit
$GISBASE/etc/GIS.sh
(if GIS.sh is not existing, it is $GISBASE/etc/Init.sh)
and add after line 2:

# set PAGER environment variable if not set
if [ ! "$PAGER" ]
then
        PAGER=more
        export PAGER
fi

Hope this helps. This bug only occurred in certain environments.

Some general comments:
> Finally, I'd like to throw out some thoughts on 5.0. I've been using
> GRASS for eight years, and just started teaching a class using 5.0. I'm
> finding the whole null value issue to be a real hassle: some frequently
> used commands do not work as expected with either nulls or zeroes,
> depending on the command. Example: r.buffer doesn't treat zero as nulls
> anymore (contrary to documentation), yet r.null doesn't work on
> reclassed files, requiring two additional steps if you want to buffer
> from a reclassed file. I'm seriously considering reverting to version
> 4.x, as the ease of use greatly surpasses the additional functionality
> of version 5.0 for me. Will these issues be addressed in the "stable"
> release, or are they inherent to the product? Thoughts, comments? If
> you're replying to this part of the message only, I guess we should
> change the subject :)
Malcolm, thanks for reporting this. The r.null problem is already
on the BUGS list, I have added the r.buffer problem.

Dear GRASS users: we need your bug reports!
Currently we are collecting all known in
  http://freegis.org/cgi-bin/viewcvs.cgi/~checkout~/grass/BUGS

The severe bugs are tagged as "release-critical" now. Means, if it
is not fixed, we can't publish GRASS 5.0stable. Like that you
can be sure to get a *stable* GRASS 5 release. The development team is 
concentrating at time on fixing these critical bugs.

Please report all bugs you know with our bug report form:
  http://www.geog.uni-hannover.de/grass/bugtracking/bugreport.html

Note: If you are using an older version than GRASS 5 beta8, check the 
NEWS file first:
  http://freegis.org/cgi-bin/viewcvs.cgi/~checkout~/grass/NEWS.html
if the bug is already fixed (this reduces mail traffic).

We encourage all users to report bugs in order to get a stable
GRASS GIS!

If you have suggestions to improve certain modules, concepts etc,
please send ot, too.

Thanks for your cooperation,

 Markus Neteler
 GRASS Development Team

-- 
Dipl.-Geogr. Markus Neteler *  University of Hannover
Institute of Physical Geography and Landscape Ecology
Schneiderberg 50 * D-30167 Hannover * Germany
Tel: ++49-(0)511-762-4494  Fax: -3984