[GRASS-dev] Re: [GRASS-SVN] r45195 - grass/branches/releasebranch_6_4/raster/r.proj.seg

GRASS GRASS developer
1

2011/1/26 <svn_grass@osgeo.org>:

Author: mmetz
Date: 2011-01-26 09:21:48 -0800 (Wed, 26 Jan 2011)
New Revision: 45195

Modified:
grass/branches/releasebranch_6_4/raster/r.proj.seg/main.c
Log:
fix for #1262

excellent catch up Markus! Would be better to modify G_message() to
allow usage G_message(NULL) for printing empty lines?

Martin

--
Martin Landa <landa.martin gmail.com> * http://geo.fsv.cvut.cz/~landa

2

On Wed, Jan 26, 2011 at 6:32 PM, Martin Landa <landa.martin@gmail.com> wrote:

2011/1/26 <svn_grass@osgeo.org>:

Author: mmetz
Date: 2011-01-26 09:21:48 -0800 (Wed, 26 Jan 2011)
New Revision: 45195

Modified:
grass/branches/releasebranch_6_4/raster/r.proj.seg/main.c
Log:
fix for #1262

excellent catch up Markus! Would be better to modify G_message() to
allow usage G_message(NULL) for printing empty lines?

IMHO no. Empty lines can be printed with G_message("") (or G_message("
") to avoid compiler warnings). G_message(NULL) occured only in
r.proj.seg (also in deactivated r.proj), but nowhere else in the code
base. The purpose of G_message() is to print a string, and "" is a
valid empty string, should be the same as "\0", whereas I regard
passing NULL as a string as pointer abuse.

Anyway, since there is no other G_message(NULL), I don't see a need to
touch G_message() or vfprint_error(). Rather modify modules than
libraries, less impact if it is only one module.

Markus M