[GRASSLIST:6267] Re: r.in.gdal LatLong projection: G_set_window(): Illegal latitude for North

[Zaboj Hrazsky, DAPHNE CZ - Institut of applied ecology]

GRASS6.0 compiled with  PROJ4 4.4.9  version runs on my debian sarge... 
but with patched  PJ_krovak.c.

Detailed description in czech ...        
http://mpa.itc.it/radim/jtsk/index.html
and here is the patch personaly...    
http://mpa.itc.it/radim/jtsk/PJ_krovak.c.patch

(shortly: run "patch PJ_krovak.c  PJ_krovak.c.patch" for the file 
PJ_krovak.c in PROJ4 src tree before compilation of PROJ4)

And everything works...

I overide the bug with Radim Blazek help:

1) to import into new location (see g.manual r.in.gdal)
2) to change projection for new location running g.setproj from xy to 
LatLong with wgs84 datum
3) to modify  "GRASSDATA_DIR"/"LOCATION"/"MAPSET"/cellhd/"MAPNAME" file 
changing first line:
from        "proj:      0"          to        "proj:        3"         
#signals type of location (xy versus LatLong)

After these r.proj was running without problems and reprojected maps 
match the others from krovak (s-jtsk) location.

... but thanks for your experiences it seems PROJ4 library need the 5 
version with  improved  PJ_krovak.c     :-).

Zaboj 

         
Kingsley Burlinson wrote:

> Zaboj
> It is possible there is a problem in the PROJ4 libraries for the 
> krovak projection and perhaps this is causing your problems. I am just 
> compiling grass now and I had major errors with PROJ4 version 4.4.9 
> and had to remove the krovak projection to get it to compile at all.
>
> Perhaps you can try another projection and see if it works in that to 
> check this potential problem.
>
> I have notified the PROJ4 list of this but have had no reply yet.
> Kingsley Burlinson   Darwin
>
> Zaboj Hrazsky wrote:
>
>> Troubles with r.in.gdal import:
>>
>> I have found following constraint when I tried import WORLDCLIM dataset:
>>
 r.in.gdal LatLong projection: G_set_window(): Illegal latitude for North