[GRASS-dev] compiling relbrand64 fails on windows

GRASS GIS GRASS GIS developer
1

tested here in the osgeo4w-environment on a WinVista-32bit-box, svn-checkout
today:

GRASS GIS compilation log
-------------------------
Started compilation: Fri Aug 31 15:48:36 GMT 2012
--
Errors in:
/osgeo4w/usr/src/grass643svn/lib/form
/osgeo4w/usr/src/grass643svn/lib/vector/diglib
/osgeo4w/usr/src/grass643svn/lib/init
[...]

syringia@NADA /osgeo4w/usr/src/grass643svn/lib/form
$ make
o /osgeo4w/usr/src/grass643svn/dist.i686-pc-msys/lib/libgrass_form.6.4.3svn
-L/osgeo4w/usr/src/grass643svn/dist.i686-pc-msys/lib -L/c/OSGeo4W/lib
-L/osgeo4w/usr/src/grass643svn/mswindows/osgeo4w/lib
OBJ.i686-pc-msys/generate.o OBJ.i686-pc-msys/open.o -lgrass_gis.6.4.3svn
-lgrass_datetime.6.4.3svn -lxdr -liberty -lws2_32 -lz -lintl
-lgrass_dbmiclient.6.4.3svn -lgrass_dbmibase.6.4.3svn -lgrass_gis.6.4.3svn
-lgrass_datetime.6.4.3svn -lxdr -liberty -lws2_32 -lz -lintl
-lgrass_gis.6.4.3svn -lgrass_datetime.6.4.3svn -lxdr -liberty -lws2_32 -lz
-lintl -lgrass_dbmibase.6.4.3svn -lgrass_gis.6.4.3svn
-lgrass_datetime.6.4.3svn -lxdr -liberty -lws2_32 -lz -lintl
../../lib/gis/OBJ.i686-pc-msys/dllmain.o
make: o: Command not found
make:
[/osgeo4w/usr/src/grass643svn/dist.i686-pc-msys/lib/libgrass_form.6.4.3svn]
Error 127 (ignored)
gcc -L/osgeo4w/usr/src/grass643svn/dist.i686-pc-msys/lib -L/c/OSGeo4W/lib
-L/osgeo4w/usr/src/grass643svn/mswindows/osgeo4w/lib -o
/osgeo4w/usr/src/grass643svn/dist.i686-pc-msys/etc/form/form
OBJ.i686-pc-msys/form.o -lgrass_dbmiclient.6.4.3svn
-lgrass_dbmibase.6.4.3svn -lgrass_gis.6.4.3svn -lgrass_datetime.6.4.3svn
-lxdr -liberty -lws2_32 -lz -lintl -lgrass_gis.6.4.3svn
-lgrass_datetime.6.4.3svn -lxdr -liberty -lws2_32 -lz -lintl
-lgrass_dbmibase.6.4.3svn -lgrass_gis.6.4.3svn -lgrass_datetime.6.4.3svn
-lxdr -liberty -lws2_32 -lz -lintl -lgrass_gis.6.4.3svn
-lgrass_datetime.6.4.3svn -lxdr -liberty -lws2_32 -lz -lintl
-lgrass_datetime.6.4.3svn \
-ltk85 -ltcl85 -lxdr -liberty -lws2_32 -lz
C:\OSGeo4W\apps\msys\bin\..\lib\gcc\mingw32\3.4.5\..\..\..\..\mingw32\bin\ld.exe:
cannot find -lgrass_dbmiclient.6.4.3svn
collect2: ld returned 1 exit status
make: *** [/osgeo4w/usr/src/grass643svn/dist.i686-pc-msys/etc/form/form]
Error 1

any idea?

-----
best regards
Helmut
--
View this message in context: http://osgeo-org.1560.n6.nabble.com/compiling-relbrand64-fails-on-windows-tp4999094.html
Sent from the Grass - Dev mailing list archive at Nabble.com.

2

On Fri, Aug 31, 2012 at 6:46 PM, Helmut Kudrnovsky <hellik@web.de> wrote:

tested here in the osgeo4w-environment on a WinVista-32bit-box, svn-checkout
today:

GRASS GIS compilation log
-------------------------
Started compilation: Fri Aug 31 15:48:36 GMT 2012
--
Errors in:
/osgeo4w/usr/src/grass643svn/lib/form
/osgeo4w/usr/src/grass643svn/lib/vector/diglib
/osgeo4w/usr/src/grass643svn/lib/init
[...]

syringia@NADA /osgeo4w/usr/src/grass643svn/lib/form
$ make
o /osgeo4w/usr/src/grass643svn/dist.i686-pc-msys/lib/libgrass_form.6.4.3svn
-L/osgeo4w/usr/src/grass643svn/dist.i686-pc-msys/lib -L/c/OSGeo4W/lib
-L/osgeo4w/usr/src/grass643svn/mswindows/osgeo4w/lib
OBJ.i686-pc-msys/generate.o OBJ.i686-pc-msys/open.o -lgrass_gis.6.4.3svn
-lgrass_datetime.6.4.3svn -lxdr -liberty -lws2_32 -lz -lintl
-lgrass_dbmiclient.6.4.3svn -lgrass_dbmibase.6.4.3svn -lgrass_gis.6.4.3svn
-lgrass_datetime.6.4.3svn -lxdr -liberty -lws2_32 -lz -lintl
-lgrass_gis.6.4.3svn -lgrass_datetime.6.4.3svn -lxdr -liberty -lws2_32 -lz
-lintl -lgrass_dbmibase.6.4.3svn -lgrass_gis.6.4.3svn
-lgrass_datetime.6.4.3svn -lxdr -liberty -lws2_32 -lz -lintl
../../lib/gis/OBJ.i686-pc-msys/dllmain.o
make: o: Command not found
make:
[/osgeo4w/usr/src/grass643svn/dist.i686-pc-msys/lib/libgrass_form.6.4.3svn]
Error 127 (ignored)
gcc -L/osgeo4w/usr/src/grass643svn/dist.i686-pc-msys/lib -L/c/OSGeo4W/lib
-L/osgeo4w/usr/src/grass643svn/mswindows/osgeo4w/lib -o
/osgeo4w/usr/src/grass643svn/dist.i686-pc-msys/etc/form/form
OBJ.i686-pc-msys/form.o -lgrass_dbmiclient.6.4.3svn
-lgrass_dbmibase.6.4.3svn -lgrass_gis.6.4.3svn -lgrass_datetime.6.4.3svn
-lxdr -liberty -lws2_32 -lz -lintl -lgrass_gis.6.4.3svn
-lgrass_datetime.6.4.3svn -lxdr -liberty -lws2_32 -lz -lintl
-lgrass_dbmibase.6.4.3svn -lgrass_gis.6.4.3svn -lgrass_datetime.6.4.3svn
-lxdr -liberty -lws2_32 -lz -lintl -lgrass_gis.6.4.3svn
-lgrass_datetime.6.4.3svn -lxdr -liberty -lws2_32 -lz -lintl
-lgrass_datetime.6.4.3svn \
-ltk85 -ltcl85 -lxdr -liberty -lws2_32 -lz
C:\OSGeo4W\apps\msys\bin\..\lib\gcc\mingw32\3.4.5\..\..\..\..\mingw32\bin\ld.exe:
cannot find -lgrass_dbmiclient.6.4.3svn
collect2: ld returned 1 exit status
make: *** [/osgeo4w/usr/src/grass643svn/dist.i686-pc-msys/etc/form/form]
Error 1

any idea?

Strange since MarkusM prepared an installer for me, also the snapshot
production works.

However, I recall
http://trac.osgeo.org/grass/ticket/469
and

"Compile errors - dllmain and fmode - releasebranch_6_4 and develbranch_6"
http://lists.osgeo.org/pipermail/grass-dev/2009-March/042676.html

Markus

3

tested here in the osgeo4w-environment on a WinVista-32bit-box, svn-checkout

today:

now tested on my Win7-64bit-box, the same also here:

GRASS GIS compilation log
-------------------------
Started compilation: Fri Aug 31 17:20:04 GMT 2012
--
Errors in:
/osgeo4w/usr/src/grass_relbranch64/lib/form
/osgeo4w/usr/src/grass_relbranch64/lib/vector/diglib
[...]

myricaria@MYRICARIA-HP /osgeo4w/usr/src/grass_relbranch64/lib/form
$ make
o
/osgeo4w/usr/src/grass_relbranch64/dist.i686-pc-mingw32/lib/libgrass_form.6.4.3svn
-L/osgeo4w/usr/src/grass_relbranch64/dist.i686-pc-mingw32/lib
-L/c/OSGeo4W/lib -L/osgeo4w/usr/src/grass_relbranch64/mswindows/osgeo4w/lib
OBJ.i686-pc-mingw32/generate.o OBJ.i686-pc-mingw32/open.o
-lgrass_gis.6.4.3svn -lgrass_datetime.6.4.3svn -lxdr -liberty -lws2_32
-lz -lintl -lgrass_dbmiclient.6.4.3svn -lgrass_dbmibase.6.4.3svn
-lgrass_gis.6.4.3svn -lgrass_datetime.6.4.3svn -lxdr -liberty -lws2_32
-lz -lintl -lgrass_gis.6.4.3svn -lgrass_datetime.6.4.3svn -lxdr -liberty
-lws2_32 -lz -lintl -lgrass_dbmibase.6.4.3svn -lgrass_gis.6.4.3svn
-lgrass_datetime.6.4.3svn -lxdr -liberty -lws2_32 -lz -lintl
../../lib/gis/OBJ.i686-pc-mingw32/dllmain.o
make: o: Command not found
make:
[/osgeo4w/usr/src/grass_relbranch64/dist.i686-pc-mingw32/lib/libgrass_form.6.4.3svn]
Error 127 (ignored)
gcc -L/osgeo4w/usr/src/grass_relbranch64/dist.i686-pc-mingw32/lib
-L/c/OSGeo4W/lib -L/osgeo4w/usr/src/grass_relbranch64/mswindows/osgeo4w/lib
-o /osgeo4w/usr/src/grass_relbranch64/dist.i686-pc-mingw32/etc/form/form
OBJ.i686-pc-mingw32/form.o -lgrass_dbmiclient.6.4.3svn
-lgrass_dbmibase.6.4.3svn -lgrass_gis.6.4.3svn -lgrass_datetime.6.4.3svn
-lxdr -liberty -lws2_32 -lz -lintl -lgrass_gis.6.4.3svn
-lgrass_datetime.6.4.3svn -lxdr -liberty -lws2_32 -lz -lintl
-lgrass_dbmibase.6.4.3svn -lgrass_gis.6.4.3svn -lgrass_datetime.6.4.3svn
-lxdr -liberty -lws2_32 -lz -lintl -lgrass_gis.6.4.3svn
-lgrass_datetime.6.4.3svn -lxdr -liberty -lws2_32 -lz -lintl
-lgrass_datetime.6.4.3svn \
                                 -ltk85 -ltcl85 -lxdr -liberty -lws2_32
-lz
C:\OSGeo4W\apps\msys\bin\..\lib\gcc\mingw32\3.4.5\..\..\..\..\mingw32\bin\ld.exe:
cannot find -lgrass_dbmiclient.6.4.3svn
collect2: ld returned 1 exit status
make: ***
[/osgeo4w/usr/src/grass_relbranch64/dist.i686-pc-mingw32/etc/form/form]
Error 1

Markus Neteler:

Strange since MarkusM prepared an installer for me, also the snapshot
production works.

maybe related to differences in the underlying osgeo4w-msys/mingw?

-----
best regards
Helmut
--
View this message in context: http://osgeo-org.1560.n6.nabble.com/compiling-relbrand64-fails-on-windows-tp4999094p4999096.html
Sent from the Grass - Dev mailing list archive at Nabble.com.

4

Helmut Kudrnovsky wrote:

tested here in the osgeo4w-environment on a WinVista-32bit-box, svn-checkout
today:

GRASS GIS compilation log
-------------------------
Started compilation: Fri Aug 31 15:48:36 GMT 2012
--
Errors in:
/osgeo4w/usr/src/grass643svn/lib/form
/osgeo4w/usr/src/grass643svn/lib/vector/diglib
/osgeo4w/usr/src/grass643svn/lib/init
[...]

syringia@NADA /osgeo4w/usr/src/grass643svn/lib/form
$ make
o /osgeo4w/usr/src/grass643svn/dist.i686-pc-msys/lib/libgrass_form.6.4.3svn

The "o" indicates that SHLIB_LD is empty, while the missing ".dll"
suffix indicates that SHLIB_SUFFIX is empty.

I have no idea why it would only affect these specific libraries; or
are you doing an incremental build (no "make clean")?

--
Glynn Clements <glynn@gclements.plus.com>

5

The "o" indicates that SHLIB_LD is empty, while the missing ".dll"
suffix indicates that SHLIB_SUFFIX is empty.

I have no idea why it would only affect these specific libraries; or
are you doing an incremental build (no "make clean")?

completely fresh svn-checkout and then following the instructions in

http://trac.osgeo.org/grass/wiki/CompileOnWindows#CompilingandInstallingGRASS6.4.x

by using

http://trac.osgeo.org/grass/browser/grass/branches/releasebranch_6_4/mswindows/osgeo4w/package.sh

maybe an (msys)/mingw/gcc-issue?

build environment based upon osgeo4w
(http://trac.osgeo.org/osgeo4w/ ,
http://download.osgeo.org/osgeo4w/versions.html )

$ gcc --version
gcc.exe (GCC) 3.4.5 (mingw-vista special r3)

-----
best regards
Helmut
--
View this message in context: http://osgeo-org.1560.n6.nabble.com/compiling-relbrand64-fails-on-windows-tp4999094p4999276.html
Sent from the Grass - Dev mailing list archive at Nabble.com.

6

$ gcc --version
gcc.exe (GCC) 3.4.5 (mingw-vista special r3)

$ make --version
GNU Make 3.81

-----
best regards
Helmut
--
View this message in context: http://osgeo-org.1560.n6.nabble.com/compiling-relbrand64-fails-on-windows-tp4999094p4999280.html
Sent from the Grass - Dev mailing list archive at Nabble.com.

7

So far everything seems to be running quite well in 6.4.3, except for some cosmetics
that I will collect and report together and the following two issues (probably not quite bugs but
a confusing behavior):

wxnviz: if a new region is set, zoom to computational region adjusts the display and nviz respects the new region, but wxnviz does not,
I had to quit GRASS with new region to get wxnviz respect it.

d.rast -o landuse96_28m cat=1,2
run in command console is interpreted as follows:
Command 'd.rast -l -a -n -d -u -s -e -9 -6 -_ -2 -8 -m
map=-o cat=1,2' failed
Details: Sorry, <l> is not a valid flag
Sorry, <a> is not a valid flag
Sorry, <n> is not a valid flag
Sorry, <d> is not a valid flag
Sorry, <u> is not a valid flag
Sorry, <s> is not a valid flag
Sorry, <e> is not a valid flag
Sorry, <9> is not a valid flag
Sorry, <6> is not a valid flag
Sorry, <_> is not a valid flag
Sorry, <2> is not a valid flag
Sorry, <8> is not a valid flag
Sorry, <m> is not a valid flag

d.rast landuse96_28m cat=1,2 -o works fine.

add volume still does not open dialog

I am running Michaels Mac binary,

Helena

Helena Mitasova
Associate Professor
Department of Marine, Earth, and Atmospheric Sciences
2800 Faucette Drive, Rm. 1125 Jordan Hall
North Carolina State University
Raleigh, NC 27695-8208
hmitaso@ncsu.edu

"All electronic mail messages in connection with State business which are sent to or received by this account are subject to the NC Public Records Law and may be disclosed to third parties.”

8

Hi,

On Wed, Sep 5, 2012 at 6:01 AM, Helena Mitasova <hmitaso@ncsu.edu> wrote:

So far everything seems to be running quite well in 6.4.3, except for some cosmetics
that I will collect and report together and the following two issues (probably not quite bugs but
a confusing behavior):

wxnviz: if a new region is set, zoom to computational region adjusts the display and nviz respects the new region, but wxnviz does not,
I had to quit GRASS with new region to get wxnviz respect it.

In grass 7 this works but not in grass 6.x because when I backported
wxNviz I wasn't able to backport also the part related to resetting
the region (G_unset_region is needed). This would need some changes in
gis library which I don't want to do, the differences between 6.x and
7 are quite large.

d.rast -o landuse96_28m cat=1,2
run in command console is interpreted as follows:
Command 'd.rast -l -a -n -d -u -s -e -9 -6 -_ -2 -8 -m
map=-o cat=1,2' failed
Details: Sorry, <l> is not a valid flag
Sorry, <a> is not a valid flag
Sorry, <n> is not a valid flag
Sorry, <d> is not a valid flag
Sorry, <u> is not a valid flag
Sorry, <s> is not a valid flag
Sorry, <e> is not a valid flag
Sorry, <9> is not a valid flag
Sorry, <6> is not a valid flag
Sorry, <_> is not a valid flag
Sorry, <2> is not a valid flag
Sorry, <8> is not a valid flag
Sorry, <m> is not a valid flag

d.rast landuse96_28m cat=1,2 -o works fine.

I tried to fix it so it should work now. I hope I didn't break
something else, this is quite difficult to test.

add volume still does not open dialog

Maybe someone else (who is more familiar with PATH and Makefiles) could help?

Anna

I am running Michaels Mac binary,

Helena

Helena Mitasova
Associate Professor
Department of Marine, Earth, and Atmospheric Sciences
2800 Faucette Drive, Rm. 1125 Jordan Hall
North Carolina State University
Raleigh, NC 27695-8208
hmitaso@ncsu.edu

"All electronic mail messages in connection with State business which are sent to or received by this account are subject to the NC Public Records Law and may be disclosed to third parties.”

_______________________________________________
grass-dev mailing list
grass-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-dev

9

On Sun, Sep 2, 2012 at 10:23 PM, Glynn Clements
<glynn@gclements.plus.com> wrote:

Helmut Kudrnovsky wrote:

tested here in the osgeo4w-environment on a WinVista-32bit-box, svn-checkout
today:

GRASS GIS compilation log
-------------------------
Started compilation: Fri Aug 31 15:48:36 GMT 2012
--
Errors in:
/osgeo4w/usr/src/grass643svn/lib/form
/osgeo4w/usr/src/grass643svn/lib/vector/diglib
/osgeo4w/usr/src/grass643svn/lib/init
[...]

I observe some differences, not sure if that's relevant here:

[neteler@north grass64]$ grep BINARY lib/gis/*
lib/gis/dllmain.c: _fmode = O_BINARY;
lib/gis/fmode.c:int _fmode = _O_BINARY;

[neteler@north grass70]$ grep BINARY lib/gis/*
lib/gis/fmode.c:int _fmode = _O_BINARY;
lib/gis/gisinit.c: _fmode = O_BINARY;
lib/gis/popen.c:#define pipe(fds) _pipe(fds, 4096, O_BINARY|O_NOINHERIT)

Markus

10

d.rast -o landuse96_28m cat=1,2
run in command console is interpreted as follows:
Command 'd.rast -l -a -n -d -u -s -e -9 -6 -_ -2 -8 -m
map=-o cat=1,2' failed
Details: Sorry, <l> is not a valid flag
Sorry, <a> is not a valid flag
Sorry, <n> is not a valid flag
Sorry, <d> is not a valid flag
Sorry, <u> is not a valid flag
Sorry, <s> is not a valid flag
Sorry, <e> is not a valid flag
Sorry, <9> is not a valid flag
Sorry, <6> is not a valid flag
Sorry, <_> is not a valid flag
Sorry, <2> is not a valid flag
Sorry, <8> is not a valid flag
Sorry, <m> is not a valid flag

d.rast landuse96_28m cat=1,2 -o works fine.

I tried to fix it so it should work now. I hope I didn't break
something else, this is quite difficult to test.

does this fix solve also the flags with d.vect? For example I get
d.vect map=-c basin_50Kval

Command 'd.vect -b -a -s -i -n -_ -5 -0 -K -v -a -l map=-c'
failed
Details: Sorry, <b> is not a valid flag
Sorry, <s> is not a valid flag
Sorry, <n> is not a valid flag
Sorry, <_> is not a valid flag
Sorry, <5> is not a valid flag
Sorry, <0> is not a valid flag
Sorry, <K> is not a valid flag
Sorry, <l> is not a valid flag

Thanks a lot.

I think I have written that the color fill on cutting planes does not work on windows - I tried it again and it works great,
We just worked in class with wxnviz last week and we did not have any problems on 20+ computers and laptops
running windows and Mac OSX, so that is fantastic, given how new the GUI is.

Also, this is the first year that I haven't seen many complaints about import/export of raster and vector data,
so that is a great improvement too.
The only comment I have seen so far was lack of Browse button when selecting the directory for v.out.ogr output,
I am wondering whether it is intentional due to some technical issues or just an oversight,
given that r.out.gdal has such option.

Helena

add volume still does not open dialog

Maybe someone else (who is more familiar with PATH and Makefiles) could help?

Anna

I am running Michaels Mac binary,

Helena

Helena Mitasova
Associate Professor
Department of Marine, Earth, and Atmospheric Sciences
2800 Faucette Drive, Rm. 1125 Jordan Hall
North Carolina State University
Raleigh, NC 27695-8208
hmitaso@ncsu.edu

"All electronic mail messages in connection with State business which are sent to or received by this account are subject to the NC Public Records Law and may be disclosed to third parties.”

_______________________________________________
grass-dev mailing list
grass-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-dev

11

On Mon, Sep 10, 2012 at 3:09 AM, Helena Mitasova <hmitaso@ncsu.edu> wrote:

d.rast -o landuse96_28m cat=1,2
run in command console is interpreted as follows:
Command 'd.rast -l -a -n -d -u -s -e -9 -6 -_ -2 -8 -m
map=-o cat=1,2' failed
Details: Sorry, <l> is not a valid flag
Sorry, <a> is not a valid flag
Sorry, <n> is not a valid flag
Sorry, <d> is not a valid flag
Sorry, <u> is not a valid flag
Sorry, <s> is not a valid flag
Sorry, <e> is not a valid flag
Sorry, <9> is not a valid flag
Sorry, <6> is not a valid flag
Sorry, <_> is not a valid flag
Sorry, <2> is not a valid flag
Sorry, <8> is not a valid flag
Sorry, <m> is not a valid flag

d.rast landuse96_28m cat=1,2 -o works fine.

I tried to fix it so it should work now. I hope I didn't break
something else, this is quite difficult to test.

does this fix solve also the flags with d.vect? For example I get
d.vect map=-c basin_50Kval

This should probably be d.vect -c map=basin_50Kval? This case works.
"d.vect map=-c basin_50Kval" is bad input and it would be difficult to
handle all wrong inputs.

Command 'd.vect -b -a -s -i -n -_ -5 -0 -K -v -a -l map=-c'
failed
Details: Sorry, <b> is not a valid flag
Sorry, <s> is not a valid flag
Sorry, <n> is not a valid flag
Sorry, <_> is not a valid flag
Sorry, <5> is not a valid flag
Sorry, <0> is not a valid flag
Sorry, <K> is not a valid flag
Sorry, <l> is not a valid flag

Thanks a lot.

I think I have written that the color fill on cutting planes does not work on windows - I tried it again and it works great,
We just worked in class with wxnviz last week and we did not have any problems on 20+ computers and laptops
running windows and Mac OSX, so that is fantastic, given how new the GUI is.

I'm glad to hear that.

Also, this is the first year that I haven't seen many complaints about import/export of raster and vector data,
so that is a great improvement too.
The only comment I have seen so far was lack of Browse button when selecting the directory for v.out.ogr output,
I am wondering whether it is intentional due to some technical issues or just an oversight,
given that r.out.gdal has such option.

I don't know, the gui respects that the dsn option is string and not
file. There is probably some difference but someone would know this
better.

Anna

Helena

add volume still does not open dialog

Maybe someone else (who is more familiar with PATH and Makefiles) could help?

Anna

I am running Michaels Mac binary,

Helena

Helena Mitasova
Associate Professor
Department of Marine, Earth, and Atmospheric Sciences
2800 Faucette Drive, Rm. 1125 Jordan Hall
North Carolina State University
Raleigh, NC 27695-8208
hmitaso@ncsu.edu

"All electronic mail messages in connection with State business which are sent to or received by this account are subject to the NC Public Records Law and may be disclosed to third parties.”

_______________________________________________
grass-dev mailing list
grass-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-dev

12

On 10/09/12 09:17, Anna Kratochvílová wrote:

On Mon, Sep 10, 2012 at 3:09 AM, Helena Mitasova<hmitaso@ncsu.edu> wrote:

Also, this is the first year that I haven't seen many complaints about import/export of raster and vector data,
so that is a great improvement too.
The only comment I have seen so far was lack of Browse button when selecting the directory for v.out.ogr output,
I am wondering whether it is intentional due to some technical issues or just an oversight,
given that r.out.gdal has such option.

I don't know, the gui respects that the dsn option is string and not
file. There is probably some difference but someone would know this
better.

ISTR that the issue is that ogr output can either be file or directory depending on the chosen format. AFAIU, it is difficult to have a browse button that allows chosing one of the two, and so the text option was the easiest answer. I don't think gdal output can be directories, or ?

But again, this is long-term memories here, so I might be wrong.

Moritz

13

On 10/09/12 11:42, Moritz Lennert wrote:

On 10/09/12 09:17, Anna Kratochvílová wrote:

On Mon, Sep 10, 2012 at 3:09 AM, Helena Mitasova<hmitaso@ncsu.edu> wrote:

Also, this is the first year that I haven't seen many complaints
about import/export of raster and vector data,
so that is a great improvement too.
The only comment I have seen so far was lack of Browse button when
selecting the directory for v.out.ogr output,
I am wondering whether it is intentional due to some technical issues
or just an oversight,
given that r.out.gdal has such option.

I don't know, the gui respects that the dsn option is string and not
file. There is probably some difference but someone would know this
better.

To complete:

ISTR that the issue is that ogr output can either be file or directory
depending on the chosen format.

Actually it can be even be more, such as database connections.

I don't think gdal output can be directories, or ?

Maybe not directories, but output to databases will be an issue once the PostGIS Raster driver in GDAL supports writing...

Moritz

14

Anna,

here is a more complete information:

this is what I typed in and running it resulted in the error message below
d.vect -c basin_50Kval

if you try to rerun it by retrieving it from history using up arrow it shows this (apparently that is what I sent without checking it)
d.vect map=-c basin_50Kval

it works with the word map
d.vect -c map=basin_50Kval

or if -c is at the end
d.vect map=basin_50Kval -c

so it unfortunately does the same thing as d.rast,

Helena

On Sep 10, 2012, at 3:17 AM, Anna Kratochvílová wrote:

On Mon, Sep 10, 2012 at 3:09 AM, Helena Mitasova <hmitaso@ncsu.edu> wrote:

d.rast -o landuse96_28m cat=1,2
run in command console is interpreted as follows:
Command 'd.rast -l -a -n -d -u -s -e -9 -6 -_ -2 -8 -m
map=-o cat=1,2' failed
Details: Sorry, <l> is not a valid flag
Sorry, <a> is not a valid flag
Sorry, <n> is not a valid flag
Sorry, <d> is not a valid flag
Sorry, <u> is not a valid flag
Sorry, <s> is not a valid flag
Sorry, <e> is not a valid flag
Sorry, <9> is not a valid flag
Sorry, <6> is not a valid flag
Sorry, <_> is not a valid flag
Sorry, <2> is not a valid flag
Sorry, <8> is not a valid flag
Sorry, <m> is not a valid flag

d.rast landuse96_28m cat=1,2 -o works fine.

I tried to fix it so it should work now. I hope I didn't break
something else, this is quite difficult to test.

does this fix solve also the flags with d.vect? For example I get
d.vect map=-c basin_50Kval

This should probably be d.vect -c map=basin_50Kval? This case works.
"d.vect map=-c basin_50Kval" is bad input and it would be difficult to
handle all wrong inputs.

Command 'd.vect -b -a -s -i -n -_ -5 -0 -K -v -a -l map=-c'
failed
Details: Sorry, <b> is not a valid flag
Sorry, <s> is not a valid flag
Sorry, <n> is not a valid flag
Sorry, <_> is not a valid flag
Sorry, <5> is not a valid flag
Sorry, <0> is not a valid flag
Sorry, <K> is not a valid flag
Sorry, <l> is not a valid flag

Thanks a lot.

I think I have written that the color fill on cutting planes does not work on windows - I tried it again and it works great,
We just worked in class with wxnviz last week and we did not have any problems on 20+ computers and laptops
running windows and Mac OSX, so that is fantastic, given how new the GUI is.

I'm glad to hear that.

Also, this is the first year that I haven't seen many complaints about import/export of raster and vector data,
so that is a great improvement too.
The only comment I have seen so far was lack of Browse button when selecting the directory for v.out.ogr output,
I am wondering whether it is intentional due to some technical issues or just an oversight,
given that r.out.gdal has such option.

I don't know, the gui respects that the dsn option is string and not
file. There is probably some difference but someone would know this
better.

Anna

Helena

add volume still does not open dialog

Maybe someone else (who is more familiar with PATH and Makefiles) could help?

Anna

I am running Michaels Mac binary,

Helena

Helena Mitasova
Associate Professor
Department of Marine, Earth, and Atmospheric Sciences
2800 Faucette Drive, Rm. 1125 Jordan Hall
North Carolina State University
Raleigh, NC 27695-8208
hmitaso@ncsu.edu

"All electronic mail messages in connection with State business which are sent to or received by this account are subject to the NC Public Records Law and may be disclosed to third parties.”

_______________________________________________
grass-dev mailing list
grass-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-dev

15

Hi,

sorry for the late answer,

On Tue, Sep 11, 2012 at 4:26 AM, Helena Mitasova <hmitaso@ncsu.edu> wrote:

Anna,

here is a more complete information:

this is what I typed in and running it resulted in the error message below
d.vect -c basin_50Kval

for me it works. I try for example

d.vect -c railroads

and the layer is added correctly and arrow up shows you

d.vect -c map=railroads@PERMANENT

There shouldn't be any platform specific code so I don't understand
why it's not working. I tested it for both grass 6.4 and 7. Is it
still an issue?

Anna

if you try to rerun it by retrieving it from history using up arrow it shows this (apparently that is what I sent without checking it)
d.vect map=-c basin_50Kval

it works with the word map
d.vect -c map=basin_50Kval

or if -c is at the end
d.vect map=basin_50Kval -c

so it unfortunately does the same thing as d.rast,

Helena

On Sep 10, 2012, at 3:17 AM, Anna Kratochvílová wrote:

On Mon, Sep 10, 2012 at 3:09 AM, Helena Mitasova <hmitaso@ncsu.edu> wrote:

d.rast -o landuse96_28m cat=1,2
run in command console is interpreted as follows:
Command 'd.rast -l -a -n -d -u -s -e -9 -6 -_ -2 -8 -m
map=-o cat=1,2' failed
Details: Sorry, <l> is not a valid flag
Sorry, <a> is not a valid flag
Sorry, <n> is not a valid flag
Sorry, <d> is not a valid flag
Sorry, <u> is not a valid flag
Sorry, <s> is not a valid flag
Sorry, <e> is not a valid flag
Sorry, <9> is not a valid flag
Sorry, <6> is not a valid flag
Sorry, <_> is not a valid flag
Sorry, <2> is not a valid flag
Sorry, <8> is not a valid flag
Sorry, <m> is not a valid flag

d.rast landuse96_28m cat=1,2 -o works fine.

I tried to fix it so it should work now. I hope I didn't break
something else, this is quite difficult to test.

does this fix solve also the flags with d.vect? For example I get
d.vect map=-c basin_50Kval

This should probably be d.vect -c map=basin_50Kval? This case works.
"d.vect map=-c basin_50Kval" is bad input and it would be difficult to
handle all wrong inputs.

Command 'd.vect -b -a -s -i -n -_ -5 -0 -K -v -a -l map=-c'
failed
Details: Sorry, <b> is not a valid flag
Sorry, <s> is not a valid flag
Sorry, <n> is not a valid flag
Sorry, <_> is not a valid flag
Sorry, <5> is not a valid flag
Sorry, <0> is not a valid flag
Sorry, <K> is not a valid flag
Sorry, <l> is not a valid flag

Thanks a lot.

I think I have written that the color fill on cutting planes does not work on windows - I tried it again and it works great,
We just worked in class with wxnviz last week and we did not have any problems on 20+ computers and laptops
running windows and Mac OSX, so that is fantastic, given how new the GUI is.

I'm glad to hear that.

Also, this is the first year that I haven't seen many complaints about import/export of raster and vector data,
so that is a great improvement too.
The only comment I have seen so far was lack of Browse button when selecting the directory for v.out.ogr output,
I am wondering whether it is intentional due to some technical issues or just an oversight,
given that r.out.gdal has such option.

I don't know, the gui respects that the dsn option is string and not
file. There is probably some difference but someone would know this
better.

Anna

Helena

add volume still does not open dialog

Maybe someone else (who is more familiar with PATH and Makefiles) could help?

Anna

I am running Michaels Mac binary,

Helena

Helena Mitasova
Associate Professor
Department of Marine, Earth, and Atmospheric Sciences
2800 Faucette Drive, Rm. 1125 Jordan Hall
North Carolina State University
Raleigh, NC 27695-8208
hmitaso@ncsu.edu

"All electronic mail messages in connection with State business which are sent to or received by this account are subject to the NC Public Records Law and may be disclosed to third parties.”

_______________________________________________
grass-dev mailing list
grass-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-dev

16

It appears that d.his does not work when the map layers are from different mapsets,
the following command which has the first map in PERMANENT and the second one in mea582f2012
expects both maps in PERMANENT
d.his i_map=elevation_shade h_map=noise bright=30
Command 'd.his bright=30 i_map=elevation_shade@PERMANENT
h_map=noise@PERMANENT' failed
Details: Raster map <noise@PERMANENT> not found

adding the mapset to the name did not help, I was still getting the error for the following:
d.his h_map=noise@mea582f2012 i_map=elevation_shade@PERMANENT bright=30
d.his i_map=elevation_shade@PERMANENT h_map=noise@mea582f2012 bright=30

It runs OK from GUI.
Can somebody try this out independently?
I can file a bug report but I would like to make sure that this is not my local problem,

Thanks, Helena

Helena Mitasova
Associate Professor
Department of Marine, Earth, and Atmospheric Sciences
2800 Faucette Drive, Rm. 1125 Jordan Hall
North Carolina State University
Raleigh, NC 27695-8208
hmitaso@ncsu.edu

"All electronic mail messages in connection with State business which are sent to or received by this account are subject to the NC Public Records Law and may be disclosed to third parties.”

On Sep 19, 2012, at 3:02 AM, Anna Kratochvílová wrote:

Hi,

sorry for the late answer,

On Tue, Sep 11, 2012 at 4:26 AM, Helena Mitasova <hmitaso@ncsu.edu> wrote:

Anna,

here is a more complete information:

this is what I typed in and running it resulted in the error message below
d.vect -c basin_50Kval

for me it works. I try for example

d.vect -c railroads

and the layer is added correctly and arrow up shows you

d.vect -c map=railroads@PERMANENT

There shouldn't be any platform specific code so I don't understand
why it's not working. I tested it for both grass 6.4 and 7. Is it
still an issue?

Anna

if you try to rerun it by retrieving it from history using up arrow it shows this (apparently that is what I sent without checking it)
d.vect map=-c basin_50Kval

it works with the word map
d.vect -c map=basin_50Kval

or if -c is at the end
d.vect map=basin_50Kval -c

so it unfortunately does the same thing as d.rast,

Helena

On Sep 10, 2012, at 3:17 AM, Anna Kratochvílová wrote:

On Mon, Sep 10, 2012 at 3:09 AM, Helena Mitasova <hmitaso@ncsu.edu> wrote:

d.rast -o landuse96_28m cat=1,2
run in command console is interpreted as follows:
Command 'd.rast -l -a -n -d -u -s -e -9 -6 -_ -2 -8 -m
map=-o cat=1,2' failed
Details: Sorry, <l> is not a valid flag
Sorry, <a> is not a valid flag
Sorry, <n> is not a valid flag
Sorry, <d> is not a valid flag
Sorry, <u> is not a valid flag
Sorry, <s> is not a valid flag
Sorry, <e> is not a valid flag
Sorry, <9> is not a valid flag
Sorry, <6> is not a valid flag
Sorry, <_> is not a valid flag
Sorry, <2> is not a valid flag
Sorry, <8> is not a valid flag
Sorry, <m> is not a valid flag

d.rast landuse96_28m cat=1,2 -o works fine.

I tried to fix it so it should work now. I hope I didn't break
something else, this is quite difficult to test.

does this fix solve also the flags with d.vect? For example I get
d.vect map=-c basin_50Kval

This should probably be d.vect -c map=basin_50Kval? This case works.
"d.vect map=-c basin_50Kval" is bad input and it would be difficult to
handle all wrong inputs.

Command 'd.vect -b -a -s -i -n -_ -5 -0 -K -v -a -l map=-c'
failed
Details: Sorry, <b> is not a valid flag
Sorry, <s> is not a valid flag
Sorry, <n> is not a valid flag
Sorry, <_> is not a valid flag
Sorry, <5> is not a valid flag
Sorry, <0> is not a valid flag
Sorry, <K> is not a valid flag
Sorry, <l> is not a valid flag

Thanks a lot.

I think I have written that the color fill on cutting planes does not work on windows - I tried it again and it works great,
We just worked in class with wxnviz last week and we did not have any problems on 20+ computers and laptops
running windows and Mac OSX, so that is fantastic, given how new the GUI is.

I'm glad to hear that.

Also, this is the first year that I haven't seen many complaints about import/export of raster and vector data,
so that is a great improvement too.
The only comment I have seen so far was lack of Browse button when selecting the directory for v.out.ogr output,
I am wondering whether it is intentional due to some technical issues or just an oversight,
given that r.out.gdal has such option.

I don't know, the gui respects that the dsn option is string and not
file. There is probably some difference but someone would know this
better.

Anna

Helena

add volume still does not open dialog

Maybe someone else (who is more familiar with PATH and Makefiles) could help?

Anna

I am running Michaels Mac binary,

Helena

Helena Mitasova
Associate Professor
Department of Marine, Earth, and Atmospheric Sciences
2800 Faucette Drive, Rm. 1125 Jordan Hall
North Carolina State University
Raleigh, NC 27695-8208
hmitaso@ncsu.edu

"All electronic mail messages in connection with State business which are sent to or received by this account are subject to the NC Public Records Law and may be disclosed to third parties.”

_______________________________________________
grass-dev mailing list
grass-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-dev

17

Hi Helena,

On Thu, Sep 20, 2012 at 5:37 AM, Helena Mitasova <hmitaso@ncsu.edu> wrote:

It appears that d.his does not work when the map layers are from different mapsets,
the following command which has the first map in PERMANENT and the second one in mea582f2012
expects both maps in PERMANENT
d.his i_map=elevation_shade h_map=noise bright=30
Command 'd.his bright=30 i_map=elevation_shade@PERMANENT
h_map=noise@PERMANENT' failed
Details: Raster map <noise@PERMANENT> not found

I hopefully fixed it in r53241. Please test also other similar
commands like d.rgb if I didn't forget about something.

adding the mapset to the name did not help, I was still getting the error for the following:
d.his h_map=noise@mea582f2012 i_map=elevation_shade@PERMANENT bright=30
d.his i_map=elevation_shade@PERMANENT h_map=noise@mea582f2012 bright=30

This should work (also before the fix). Maybe you could get confused
because the error message appears again after re-rendering. To avoid
this you must remove the layer or double click to open properties
dialog and fix it there.

It runs OK from GUI.
Can somebody try this out independently?
I can file a bug report but I would like to make sure that this is not my local problem,

Thanks, Helena

Best,
Anna

Helena Mitasova
Associate Professor
Department of Marine, Earth, and Atmospheric Sciences
2800 Faucette Drive, Rm. 1125 Jordan Hall
North Carolina State University
Raleigh, NC 27695-8208
hmitaso@ncsu.edu

"All electronic mail messages in connection with State business which are sent to or received by this account are subject to the NC Public Records Law and may be disclosed to third parties.”

On Sep 19, 2012, at 3:02 AM, Anna Kratochvílová wrote:

Hi,

sorry for the late answer,

On Tue, Sep 11, 2012 at 4:26 AM, Helena Mitasova <hmitaso@ncsu.edu> wrote:

Anna,

here is a more complete information:

this is what I typed in and running it resulted in the error message below
d.vect -c basin_50Kval

for me it works. I try for example

d.vect -c railroads

and the layer is added correctly and arrow up shows you

d.vect -c map=railroads@PERMANENT

There shouldn't be any platform specific code so I don't understand
why it's not working. I tested it for both grass 6.4 and 7. Is it
still an issue?

Anna

if you try to rerun it by retrieving it from history using up arrow it shows this (apparently that is what I sent without checking it)
d.vect map=-c basin_50Kval

it works with the word map
d.vect -c map=basin_50Kval

or if -c is at the end
d.vect map=basin_50Kval -c

so it unfortunately does the same thing as d.rast,

Helena

On Sep 10, 2012, at 3:17 AM, Anna Kratochvílová wrote:

On Mon, Sep 10, 2012 at 3:09 AM, Helena Mitasova <hmitaso@ncsu.edu> wrote:

d.rast -o landuse96_28m cat=1,2
run in command console is interpreted as follows:
Command 'd.rast -l -a -n -d -u -s -e -9 -6 -_ -2 -8 -m
map=-o cat=1,2' failed
Details: Sorry, <l> is not a valid flag
Sorry, <a> is not a valid flag
Sorry, <n> is not a valid flag
Sorry, <d> is not a valid flag
Sorry, <u> is not a valid flag
Sorry, <s> is not a valid flag
Sorry, <e> is not a valid flag
Sorry, <9> is not a valid flag
Sorry, <6> is not a valid flag
Sorry, <_> is not a valid flag
Sorry, <2> is not a valid flag
Sorry, <8> is not a valid flag
Sorry, <m> is not a valid flag

d.rast landuse96_28m cat=1,2 -o works fine.

I tried to fix it so it should work now. I hope I didn't break
something else, this is quite difficult to test.

does this fix solve also the flags with d.vect? For example I get
d.vect map=-c basin_50Kval

This should probably be d.vect -c map=basin_50Kval? This case works.
"d.vect map=-c basin_50Kval" is bad input and it would be difficult to
handle all wrong inputs.

Command 'd.vect -b -a -s -i -n -_ -5 -0 -K -v -a -l map=-c'
failed
Details: Sorry, <b> is not a valid flag
Sorry, <s> is not a valid flag
Sorry, <n> is not a valid flag
Sorry, <_> is not a valid flag
Sorry, <5> is not a valid flag
Sorry, <0> is not a valid flag
Sorry, <K> is not a valid flag
Sorry, <l> is not a valid flag

Thanks a lot.

I think I have written that the color fill on cutting planes does not work on windows - I tried it again and it works great,
We just worked in class with wxnviz last week and we did not have any problems on 20+ computers and laptops
running windows and Mac OSX, so that is fantastic, given how new the GUI is.

I'm glad to hear that.

Also, this is the first year that I haven't seen many complaints about import/export of raster and vector data,
so that is a great improvement too.
The only comment I have seen so far was lack of Browse button when selecting the directory for v.out.ogr output,
I am wondering whether it is intentional due to some technical issues or just an oversight,
given that r.out.gdal has such option.

I don't know, the gui respects that the dsn option is string and not
file. There is probably some difference but someone would know this
better.

Anna

Helena

add volume still does not open dialog

Maybe someone else (who is more familiar with PATH and Makefiles) could help?

Anna

I am running Michaels Mac binary,

Helena

Helena Mitasova
Associate Professor
Department of Marine, Earth, and Atmospheric Sciences
2800 Faucette Drive, Rm. 1125 Jordan Hall
North Carolina State University
Raleigh, NC 27695-8208
hmitaso@ncsu.edu

"All electronic mail messages in connection with State business which are sent to or received by this account are subject to the NC Public Records Law and may be disclosed to third parties.”

_______________________________________________
grass-dev mailing list
grass-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-dev