[GRASS-dev] Re: [GRASS GIS] #1110: v.rast.stats locks up on wingrass

[sonja]

GRASS GIS <trac <at> osgeo.org> writes:

> 
> #1110: v.rast.stats locks up on wingrass
> ---------------------------------------------------------+------------------
>  Reporter:  hamish                                       |       Owner: 
grass-dev@…              
>      Type:  defect                                       |      Status:  new 
                    
>  Priority:  normal                                       |   Milestone:  6.4.0
                   
> Component:  Vector                                       |     Version: 
svn-releasebranch64      
>  Keywords:  wingrass, v.rast.stats, v.db.addcol, v.info  |    Platform: 
MSWindows XP             
>       Cpu:  x86-32                                       |  
> ---------------------------------------------------------+------------------
> 
> Comment(by hamish):
> 
>  Yvan wrote:
>  {{{
>  This is the function and the message :
> 
>  (Mon Jul 19 15:12:58 2010)
>  v.rast.stats vector=parcelles_masquees <at> mercier raster=mntyzeron_5m <at>
mercier
>  colprefix=mnt
>  C:/GRASS-64-SVN/scripts/v.rast.stats: line 268: v.db.addcol:
>  command not found
>  ERROR: Cannot continue (problem adding columns).
>  (Mon Jul 19 15:13:06 2010) Command finished (7 sec)
>  }}}
> 
> 
>  although I've only seen the subprocess freezing manifestation of this.
> 

I get the same error using Grass 6.4 and 6.5.
Might this be related to the fact, that v.db.addcol is a shell script, the same
as v.rast.stats? 
As far as I understand, while starting a shell script under windows, GRASS_SH
(which is set to sh.exe in the environment variables) will start the .bat file
in the /bin directory. This .bat file points to the shell script in the /scripts
directory. 
However, once sh.exe started to read the shell script (here v.rast.stats), in
which directory will it search a second shell script referenced in the first
shell script (here v.db.addcol)? In the same?
Is there anything to fix this bug?
Thanks,
Sonja