[GRASS-dev] grass7 on mac OSX

GRASS GRASS developer
1

On May 24, 2010, at 6:35 AM, grass-dev-request@lists.osgeo.org wrote:

Date: Mon, 24 May 2010 00:49:51 +0200
From: Massimo Di Stefano <massimodisasha@gmail.com>
Subject: [GRASS-dev] grass7 on mac OSX
To: grass-dev list <grass-dev@lists.osgeo.org>
Message-ID: <05BC922B-8790-4579-93E3-3052D5B1518D@gmail.com>
Content-Type: text/plain; charset=us-ascii

Hi All,

i upgrade grass7 from svn on mac osx 10.6.3 ,
the build ends with an error in wxpython, rerunning make inside the wxpython director fix the problem.

I've reported this too. There is something wrong in the build/make files somewhere.

but tring to load grass from python i had thi error :

In [1]: from grass.lib import grass
------------------------------------------------------------
Traceback (most recent call last):
File "<ipython console>", line 1, in <module>
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/python/grass/lib/__init__.py", line 3, in <module>
   # $Id$
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/python/grass/lib/math.py", line 7, in <module>
ImportError: dlopen(/Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so, 2): Symbol not found: _G_math_backward_solving
Referenced from: /Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so
Expected in: /Applications/GRASS-7.0.app/Contents/MacOS/lib/libgrass_gmath.dylib
in /Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so

have you any clue on what i'm missing ?
thanks!

Massimo.

I think this is a bogus error. The identical thing happened to a student here except that it claimed to be missing the subprocess module. The module is not missing and GRASS 6.5 works fine. GRASS 7 runs fine on my Mac, but this kind of error is cropping up somewhat randomly apparently.

Michael

2

Il giorno 24/mag/2010, alle ore 17.28, Michael Barton ha scritto:

On May 24, 2010, at 6:35 AM, grass-dev-request@lists.osgeo.org wrote:

Date: Mon, 24 May 2010 00:49:51 +0200
From: Massimo Di Stefano <massimodisasha@gmail.com>
Subject: [GRASS-dev] grass7 on mac OSX
To: grass-dev list <grass-dev@lists.osgeo.org>
Message-ID: <05BC922B-8790-4579-93E3-3052D5B1518D@gmail.com>
Content-Type: text/plain; charset=us-ascii

Hi All,

i upgrade grass7 from svn on mac osx 10.6.3 ,
the build ends with an error in wxpython, rerunning make inside the wxpython director fix the problem.

I've reported this too. There is something wrong in the build/make files somewhere.

but tring to load grass from python i had thi error :

In [1]: from grass.lib import grass
------------------------------------------------------------
Traceback (most recent call last):
File "<ipython console>", line 1, in <module>
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/python/grass/lib/__init__.py", line 3, in <module>
  # $Id$
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/python/grass/lib/math.py", line 7, in <module>
ImportError: dlopen(/Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so, 2): Symbol not found: _G_math_backward_solving
Referenced from: /Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so
Expected in: /Applications/GRASS-7.0.app/Contents/MacOS/lib/libgrass_gmath.dylib
in /Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so

have you any clue on what i'm missing ?
thanks!

Massimo.

I think this is a bogus error. The identical thing happened to a student here except that it claimed to be missing the subprocess module. The module is not missing and GRASS 6.5 works fine. GRASS 7 runs fine on my Mac, but this kind of error is cropping up somewhat randomly apparently.

Michael

Hi Michael, so it works for you, good :slight_smile:
have you any clue on what kind of changes have you did to get it running ?

i'm using :

Python2.6.1 ( standard python from system, osx snow leopard )

Ctypesgen from svn :
Repository Root: http://ctypesgen.googlecode.com/svn
Revision: 107

Thanks for any help,

Massimo

3

On May 26, 2010, at 7:38 PM, Massimo Di Stefano wrote:

Il giorno 24/mag/2010, alle ore 17.28, Michael Barton ha scritto:

On May 24, 2010, at 6:35 AM, grass-dev-request@lists.osgeo.org wrote:

Date: Mon, 24 May 2010 00:49:51 +0200
From: Massimo Di Stefano <massimodisasha@gmail.com>
Subject: [GRASS-dev] grass7 on mac OSX
To: grass-dev list <grass-dev@lists.osgeo.org>
Message-ID: <05BC922B-8790-4579-93E3-3052D5B1518D@gmail.com>
Content-Type: text/plain; charset=us-ascii

Hi All,

i upgrade grass7 from svn on mac osx 10.6.3 ,
the build ends with an error in wxpython, rerunning make inside the wxpython director fix the problem.

I've reported this too. There is something wrong in the build/make files somewhere.

but tring to load grass from python i had thi error :

In [1]: from grass.lib import grass
------------------------------------------------------------
Traceback (most recent call last):
File "<ipython console>", line 1, in <module>
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/python/grass/lib/__init__.py", line 3, in <module>
# $Id$
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/python/grass/lib/math.py", line 7, in <module>
ImportError: dlopen(/Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so, 2): Symbol not found: _G_math_backward_solving
Referenced from: /Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so
Expected in: /Applications/GRASS-7.0.app/Contents/MacOS/lib/libgrass_gmath.dylib
in /Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so

have you any clue on what i'm missing ?
thanks!

Massimo.

I think this is a bogus error. The identical thing happened to a student here except that it claimed to be missing the subprocess module. The module is not missing and GRASS 6.5 works fine. GRASS 7 runs fine on my Mac, but this kind of error is cropping up somewhat randomly apparently.

Michael

Hi Michael, so it works for you, good :slight_smile:
have you any clue on what kind of changes have you did to get it running ?

i'm using :

Python2.6.1 ( standard python from system, osx snow leopard )

Ctypesgen from svn :
Repository Root: http://ctypesgen.googlecode.com/svn
Revision: 107

Thanks for any help,

Massimo

I did nothing special to make it work. I'm also using the standard system python 2.6.1. I did a build and it worked fine. I posted my binaries for students and colleagues. A student downloaded it and it won't run. She has a very similar machine.

Sorry I'm no help. I think it is something else wrong and that this error is probably not what is really wrong.

Michael

4

i see ctypesgen is included in grass7 (not ctype itself)
to check if my local copy of ctypesgen is the reason of the error,
i temporary moved ctypesgen from my computer :

sudo mv /Library/Python/2.6/site-packages/ctypesgencore /Library/Python/2.6/site-packages/ctypesgencore__
sudo mv /usr/local/bin/ctypesgen.py /usr/local/bin/ctypesgen__.py

i can't remove ctype, is is used by other apps

then in the grass_trunk source :

make distclean
svn up
./configure ...
make

but the problem persists :

In [2]: import os, sys

In [3]: from grass.lib import raster
------------------------------------------------------------
Traceback (most recent call last):
  File "<ipython console>", line 1, in <module>
  File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/python/grass/lib/__init__.py", line 3, in <module>
    # $Id$
  File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/python/grass/lib/math.py", line 7, in <module>
ImportError: dlopen(/Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so, 2): Symbol not found: _G_math_backward_solving
  Referenced from: /Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so
  Expected in: /Applications/GRASS-7.0.app/Contents/MacOS/lib/libgrass_gmath.dylib
in /Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so

Massimo.

Il giorno 27/mag/2010, alle ore 06.28, Markus Neteler ha scritto:

I thought that ctypes are now *included*?
Perhaps it is now conflicting for you if you have two copies?

Markus

Michael Barton wrote:

Yes. I installed ctypes. No problem there. He said that users of the binary I compiled do not have to have ctypes installed separately. But I wonder if it has something to do with this.

Michael
____________________
C. Michael Barton
Director, Center for Social Dynamics& Complexity
Professor of Anthropology, School of Human Evolution& Social Change
Arizona State University

voice: 480-965-6262 (SHESC), 480-727-9746 (CSDC)
fax: 480-965-7671 (SHESC), 480-727-0709 (CSDC)
www: www.public.asu.edu/~cmbarton, http://csdc.asu.edu

On May 26, 2010, at 9:11 PM, Markus Neteler wrote:

On Thu, May 27, 2010 at 6:07 AM, Michael Barton<Michael.Barton@asu.edu> wrote:
...
    

I did nothing special to make it work. I'm also using the standard system python 2.6.1. I did a build and it worked fine. I posted my binaries for students and colleagues. A student downloaded it and it won't run. She has a very similar machine.

Sorry I'm no help. I think it is something else wrong and that this error is probably not what is really wrong.
      

Do you/they use latest GRASS 7? Glynn included a Ctypes* copy in SVN, this
won't still work?

Markus
    

5

Hello Massimo,
can you please make a svn update?The gmath function
G_math_backward_solving has been renamed into
G_math_backward_substitution a while ago. Maybe something got mixed
up?

Best
Soeren

2010/5/27 Massimo Di Stefano <massimodisasha@gmail.com>:

i see ctypesgen is included in grass7 (not ctype itself)
to check if my local copy of ctypesgen is the reason of the error,
i temporary moved ctypesgen from my computer :

sudo mv /Library/Python/2.6/site-packages/ctypesgencore /Library/Python/2.6/site-packages/ctypesgencore__
sudo mv /usr/local/bin/ctypesgen.py /usr/local/bin/ctypesgen__.py

i can't remove ctype, is is used by other apps

then in the grass_trunk source :

make distclean
svn up
./configure ...
make

but the problem persists :

In [2]: import os, sys

In [3]: from grass.lib import raster
------------------------------------------------------------
Traceback (most recent call last):
File "<ipython console>", line 1, in <module>
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/python/grass/lib/__init__.py", line 3, in <module>
# $Id$
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/python/grass/lib/math.py", line 7, in <module>
ImportError: dlopen(/Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so, 2): Symbol not found: _G_math_backward_solving
Referenced from: /Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so
Expected in: /Applications/GRASS-7.0.app/Contents/MacOS/lib/libgrass_gmath.dylib
in /Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so

Massimo.

Il giorno 27/mag/2010, alle ore 06.28, Markus Neteler ha scritto:

I thought that ctypes are now *included*?
Perhaps it is now conflicting for you if you have two copies?

Markus

Michael Barton wrote:

Yes. I installed ctypes. No problem there. He said that users of the binary I compiled do not have to have ctypes installed separately. But I wonder if it has something to do with this.

Michael
____________________
C. Michael Barton
Director, Center for Social Dynamics& Complexity
Professor of Anthropology, School of Human Evolution& Social Change
Arizona State University

voice: 480-965-6262 (SHESC), 480-727-9746 (CSDC)
fax: 480-965-7671 (SHESC), 480-727-0709 (CSDC)
www: www.public.asu.edu/~cmbarton, http://csdc.asu.edu

On May 26, 2010, at 9:11 PM, Markus Neteler wrote:

On Thu, May 27, 2010 at 6:07 AM, Michael Barton<Michael.Barton@asu.edu> wrote:
...

I did nothing special to make it work. I'm also using the standard system python 2.6.1. I did a build and it worked fine. I posted my binaries for students and colleagues. A student downloaded it and it won't run. She has a very similar machine.

Sorry I'm no help. I think it is something else wrong and that this error is probably not what is really wrong.

Do you/they use latest GRASS 7? Glynn included a Ctypes* copy in SVN, this
won't still work?

Markus

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

6

Hell Soeren,

seems i've an up-to-date revision :

MacBook-Pro-15-di-Massimo-Di-Stefano:grass_trunk sasha$ svn up
At revision 42358.

i also did make distclean first to re-try configure && make.

but your note about a "renamed" function, point me to a problem i had on qgis side too (make install overwrite the .app but ... it don't do it as aspected)

infact, removing manually the old grass70.app (compiled from source) trashing the complete app from /Application , instead of overwrite it using "sudo make install"

now i have :

In [1]: import os, sys

In [2]: from grass.lib import raster
------------------------------------------------------------
Traceback (most recent call last):
  File "<ipython console>", line 1, in <module>
ImportError: No module named lib

the error is present but changes ... this time maybe can be a wrong sintax i used ???
weird .. i'm a bit confused :wink:

i'm tring to replicate, from command line, the code :

###

import os, sys
from grass.lib import grass, raster
from ctypes import *

input = sys.argv[1]

grass.G_gisinit(sys.argv[0])

mapset = grass.G_find_raster2(input, "")
print mapset

Massimo

Il giorno 27/mag/2010, alle ore 10.37, Soeren Gebbert ha scritto:

Hello Massimo,
can you please make a svn update?The gmath function
G_math_backward_solving has been renamed into
G_math_backward_substitution a while ago. Maybe something got mixed
up?

Best
Soeren

2010/5/27 Massimo Di Stefano <massimodisasha@gmail.com>:

i see ctypesgen is included in grass7 (not ctype itself)
to check if my local copy of ctypesgen is the reason of the error,
i temporary moved ctypesgen from my computer :

sudo mv /Library/Python/2.6/site-packages/ctypesgencore /Library/Python/2.6/site-packages/ctypesgencore__
sudo mv /usr/local/bin/ctypesgen.py /usr/local/bin/ctypesgen__.py

i can't remove ctype, is is used by other apps

then in the grass_trunk source :

make distclean
svn up
./configure ...
make

but the problem persists :

In [2]: import os, sys

In [3]: from grass.lib import raster
------------------------------------------------------------
Traceback (most recent call last):
File "<ipython console>", line 1, in <module>
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/python/grass/lib/__init__.py", line 3, in <module>
   # $Id$
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/python/grass/lib/math.py", line 7, in <module>
ImportError: dlopen(/Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so, 2): Symbol not found: _G_math_backward_solving
Referenced from: /Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so
Expected in: /Applications/GRASS-7.0.app/Contents/MacOS/lib/libgrass_gmath.dylib
in /Applications/GRASS-7.0.app/Contents/MacOS/etc/python/grass/lib/_math.so

Massimo.

Il giorno 27/mag/2010, alle ore 06.28, Markus Neteler ha scritto:

I thought that ctypes are now *included*?
Perhaps it is now conflicting for you if you have two copies?

Markus

Michael Barton wrote:

Yes. I installed ctypes. No problem there. He said that users of the binary I compiled do not have to have ctypes installed separately. But I wonder if it has something to do with this.

Michael
____________________
C. Michael Barton
Director, Center for Social Dynamics& Complexity
Professor of Anthropology, School of Human Evolution& Social Change
Arizona State University

voice: 480-965-6262 (SHESC), 480-727-9746 (CSDC)
fax: 480-965-7671 (SHESC), 480-727-0709 (CSDC)
www: www.public.asu.edu/~cmbarton, http://csdc.asu.edu

On May 26, 2010, at 9:11 PM, Markus Neteler wrote:

On Thu, May 27, 2010 at 6:07 AM, Michael Barton<Michael.Barton@asu.edu> wrote:
...

I did nothing special to make it work. I'm also using the standard system python 2.6.1. I did a build and it worked fine. I posted my binaries for students and colleagues. A student downloaded it and it won't run. She has a very similar machine.

Sorry I'm no help. I think it is something else wrong and that this error is probably not what is really wrong.

Do you/they use latest GRASS 7? Glynn included a Ctypes* copy in SVN, this
won't still work?

Markus

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

7

On May 27, 2010, at 3:56 AM, Massimo Di Stefano <massimodisasha@gmail.com > wrote:

Hell Soeren,

seems i've an up-to-date revision :

MacBook-Pro-15-di-Massimo-Di-Stefano:grass_trunk sasha$ svn up
At revision 42358.

i also did make distclean first to re-try configure && make.

but your note about a "renamed" function, point me to a problem i had on qgis side too (make install overwrite the .app but ... it don't do it as aspected)

infact, removing manually the old grass70.app (compiled from source) trashing the complete app from /Application , instead of overwrite it using "sudo make install"

now i have :

In [1]: import os, sys

In [2]: from grass.lib import raster
------------------------------------------------------------
Traceback (most recent call last):
File "<ipython console>", line 1, in <module>
ImportError: No module named lib

I wonder if ipython is affecting this? Just a random idea. (I can't try anything right now- dev Mac died)

the error is present but changes ... this time maybe can be a wrong sintax i used ???
weird .. i'm a bit confused :wink:

i'm tring to replicate, from command line, the code :

###

import os, sys
from grass.lib import grass, raster
from ctypes import *

input = sys.argv[1]

grass.G_gisinit(sys.argv[0])

mapset = grass.G_find_raster2(input, "")
print mapset

Massimo

Il giorno 27/mag/2010, alle ore 10.37, Soeren Gebbert ha scritto:

Hello Massimo,
can you please make a svn update?The gmath function
G_math_backward_solving has been renamed into
G_math_backward_substitution a while ago. Maybe something got mixed
up?

Best
Soeren

8

i tried a new build, removing the old .app first
i tried without ipython, but the problem is the same :

GRASS 7.0.svn (spearfish60):~ > python
Python 2.6.1 (r261:67515, Feb 11 2010, 00:51:29)
[GCC 4.2.1 (Apple Inc. build 5646)] on darwin
Type "help", "copyright", "credits" or "license" for more information.

import os, sys
from grass.lib import raster

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ImportError: No module named lib

Il giorno 27/mag/2010, alle ore 15.43, William Kyngesburye ha scritto:

On May 27, 2010, at 3:56 AM, Massimo Di Stefano <massimodisasha@gmail.com> wrote:

Hell Soeren,

seems i've an up-to-date revision :

MacBook-Pro-15-di-Massimo-Di-Stefano:grass_trunk sasha$ svn up
At revision 42358.

i also did make distclean first to re-try configure && make.

but your note about a "renamed" function, point me to a problem i had on qgis side too (make install overwrite the .app but ... it don't do it as aspected)

infact, removing manually the old grass70.app (compiled from source) trashing the complete app from /Application , instead of overwrite it using "sudo make install"

now i have :

In [1]: import os, sys

In [2]: from grass.lib import raster
------------------------------------------------------------
Traceback (most recent call last):
File "<ipython console>", line 1, in <module>
ImportError: No module named lib

I wonder if ipython is affecting this? Just a random idea. (I can't try anything right now- dev Mac died)

the error is present but changes ... this time maybe can be a wrong sintax i used ???
weird .. i'm a bit confused :wink:

i'm tring to replicate, from command line, the code :

###

import os, sys
from grass.lib import grass, raster
from ctypes import *

input = sys.argv[1]

grass.G_gisinit(sys.argv[0])

mapset = grass.G_find_raster2(input, "")
print mapset

Massimo

Il giorno 27/mag/2010, alle ore 10.37, Soeren Gebbert ha scritto:

Hello Massimo,
can you please make a svn update?The gmath function
G_math_backward_solving has been renamed into
G_math_backward_substitution a while ago. Maybe something got mixed
up?

Best
Soeren

9

No ipython on the student's machine.

I compiled GRASS 7 and had no problems. I posted my binary so others could use it. The one person who has tried it gets a bogus error about Python subprocess() not existing. The binary of GRASS 6.5 that I compiled and posted at the same time runs fine on her machine (demonstrating that subprocess() is OK).

Oddly, some other folks here got similar bogus errors with GRASS 6.5 or 7 (I can't remember) when trying to run a Python script under winGRASS a few weeks back.

Michael
____________________
C. Michael Barton
Director, Center for Social Dynamics & Complexity
Professor of Anthropology, School of Human Evolution & Social Change
Arizona State University

voice: 480-965-6262 (SHESC), 480-727-9746 (CSDC)
fax: 480-965-7671 (SHESC), 480-727-0709 (CSDC)
www: www.public.asu.edu/~cmbarton, http://csdc.asu.edu

On May 27, 2010, at 6:43 AM, William Kyngesburye wrote:

On May 27, 2010, at 3:56 AM, Massimo Di Stefano <massimodisasha@gmail.com

wrote:

Hell Soeren,

seems i've an up-to-date revision :

MacBook-Pro-15-di-Massimo-Di-Stefano:grass_trunk sasha$ svn up
At revision 42358.

i also did make distclean first to re-try configure && make.

but your note about a "renamed" function, point me to a problem i
had on qgis side too (make install overwrite the .app but ... it
don't do it as aspected)

infact, removing manually the old grass70.app (compiled from source)
trashing the complete app from /Application , instead of overwrite
it using "sudo make install"

now i have :

In [1]: import os, sys

In [2]: from grass.lib import raster
------------------------------------------------------------
Traceback (most recent call last):
File "<ipython console>", line 1, in <module>
ImportError: No module named lib

I wonder if ipython is affecting this? Just a random idea. (I can't
try anything right now- dev Mac died)

the error is present but changes ... this time maybe can be a wrong
sintax i used ???
weird .. i'm a bit confused :wink:

i'm tring to replicate, from command line, the code :

###

import os, sys
from grass.lib import grass, raster
from ctypes import *

input = sys.argv[1]

grass.G_gisinit(sys.argv[0])

mapset = grass.G_find_raster2(input, "")
print mapset

Massimo

Il giorno 27/mag/2010, alle ore 10.37, Soeren Gebbert ha scritto:

Hello Massimo,
can you please make a svn update?The gmath function
G_math_backward_solving has been renamed into
G_math_backward_substitution a while ago. Maybe something got mixed
up?

Best
Soeren

10

Now the error i means :

"""

In [2]: from grass.lib import raster
------------------------------------------------------------
Traceback (most recent call last):
File "<ipython console>", line 1, in <module>
ImportError: No module named lib

"""

is reflected in the gui.
the new GUI code changes are affected by it,
tring to build grass70, starting the application
i have a broken gui that show me this error log :

GRASS 7.0.svn (spearfish60):~ >
WARNING: Vector digitizer is not available (dlopen(/Applications/GRASS-7.0.app/Contents/MacOS/etc/gui/wxpython/vdigit/_grass7_wxvdigit.so, 2): Symbol not found: __Py_RefTotal
  Referenced from: /Applications/GRASS-7.0.app/Contents/MacOS/etc/gui/wxpython/vdigit/_grass7_wxvdigit.so
  Expected in: flat namespace
in /Applications/GRASS-7.0.app/Contents/MacOS/etc/gui/wxpython/vdigit/_grass7_wxvdigit.so).

Note that the vector digitizer is currently not working under MS Windows (hopefully this will be fixed soon). Please keep an eye out for updated versions of GRASS.
Traceback (most recent call last):
  File "/Applications/GRASS-7.0.app/Contents/MacOS/etc/gui/wxpython/wxgui.py", line 87, in <module>
    import gui_modules.nviz_tools as nviz_tools
  File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/nviz_tools.py", line 36, in <module>
    from nviz_mapdisp import wxUpdateView as wxUpdateView
  File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/nviz_mapdisp.py", line 41, in <module>
    import wxnviz
  File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/wxnviz.py", line 24, in <module>
    from grass.lib.grass import *
ImportError: No module named lib.grass

tring to re-run : "g.gui" i have the same log.

just to check if my python "is bugged"
i build python2.6.5 source code,
in a non standard location "/usr/local/gislib/unix",
as unix library (no framework)
then i exported the path for the new python
and i used it to reconfigure-compile-install grass70,
but unluky i have problems to build wxpython
(i compiled wxwidjet as 32+64 bit but i'm having
problems to built wxpython on it)
i'll try to figure out a wxpython cocoa build
and tell you if the grass.lib not found problem persists.

thanks for any help.

Il giorno 27/mag/2010, alle ore 16.52, Michael Barton ha scritto:

No ipython on the student's machine.

I compiled GRASS 7 and had no problems. I posted my binary so others could use it. The one person who has tried it gets a bogus error about Python subprocess() not existing. The binary of GRASS 6.5 that I compiled and posted at the same time runs fine on her machine (demonstrating that subprocess() is OK).

Oddly, some other folks here got similar bogus errors with GRASS 6.5 or 7 (I can't remember) when trying to run a Python script under winGRASS a few weeks back.

Michael
____________________
C. Michael Barton
Director, Center for Social Dynamics & Complexity
Professor of Anthropology, School of Human Evolution & Social Change
Arizona State University

voice: 480-965-6262 (SHESC), 480-727-9746 (CSDC)
fax: 480-965-7671 (SHESC), 480-727-0709 (CSDC)
www: www.public.asu.edu/~cmbarton, http://csdc.asu.edu

On May 27, 2010, at 6:43 AM, William Kyngesburye wrote:

On May 27, 2010, at 3:56 AM, Massimo Di Stefano <massimodisasha@gmail.com

wrote:

Hell Soeren,

seems i've an up-to-date revision :

MacBook-Pro-15-di-Massimo-Di-Stefano:grass_trunk sasha$ svn up
At revision 42358.

i also did make distclean first to re-try configure && make.

but your note about a "renamed" function, point me to a problem i
had on qgis side too (make install overwrite the .app but ... it
don't do it as aspected)

infact, removing manually the old grass70.app (compiled from source)
trashing the complete app from /Application , instead of overwrite
it using "sudo make install"

now i have :

In [1]: import os, sys

In [2]: from grass.lib import raster
------------------------------------------------------------
Traceback (most recent call last):
File "<ipython console>", line 1, in <module>
ImportError: No module named lib

I wonder if ipython is affecting this? Just a random idea. (I can't
try anything right now- dev Mac died)

the error is present but changes ... this time maybe can be a wrong
sintax i used ???
weird .. i'm a bit confused :wink:

i'm tring to replicate, from command line, the code :

###

import os, sys
from grass.lib import grass, raster
from ctypes import *

input = sys.argv[1]

grass.G_gisinit(sys.argv[0])

mapset = grass.G_find_raster2(input, "")
print mapset

Massimo

Il giorno 27/mag/2010, alle ore 10.37, Soeren Gebbert ha scritto:

Hello Massimo,
can you please make a svn update?The gmath function
G_math_backward_solving has been renamed into
G_math_backward_substitution a while ago. Maybe something got mixed
up?

Best
Soeren

11

I just compiled GRASS 7 a couple days ago, before leaving for Spain. Wx digitizer and nviz don't compile. Also others can't even open my binaries.

Is there a new configure switch for ctypes? Are we supposed to no longer use the python switch?

Michael
____________________
C. Michael Barton
Director, Center for Social Dynamics & Complexity
Professor of Anthropology, School of Human Evolution & Social Change
Arizona State University

voice: 480-965-6262 (SHESC), 480-727-9746 (CSDC)
fax: 480-965-7671 (SHESC), 480-727-0709 (CSDC)
www: www.public.asu.edu/~cmbarton, http://csdc.asu.edu

On Jun 1, 2010, at 11:36 PM, Massimo Di Stefano wrote:

Now the error i means :

"""

In [2]: from grass.lib import raster
------------------------------------------------------------
Traceback (most recent call last):
File "<ipython console>", line 1, in <module>
ImportError: No module named lib

"""

is reflected in the gui.
the new GUI code changes are affected by it,
tring to build grass70, starting the application
i have a broken gui that show me this error log :

GRASS 7.0.svn (spearfish60):~ >
WARNING: Vector digitizer is not available (dlopen(/Applications/GRASS-7.0.app/Contents/MacOS/etc/gui/wxpython/vdigit/_grass7_wxvdigit.so, 2): Symbol not found: __Py_RefTotal
Referenced from: /Applications/GRASS-7.0.app/Contents/MacOS/etc/gui/wxpython/vdigit/_grass7_wxvdigit.so
Expected in: flat namespace
in /Applications/GRASS-7.0.app/Contents/MacOS/etc/gui/wxpython/vdigit/_grass7_wxvdigit.so).

Note that the vector digitizer is currently not working under MS Windows (hopefully this will be fixed soon). Please keep an eye out for updated versions of GRASS.
Traceback (most recent call last):
File "/Applications/GRASS-7.0.app/Contents/MacOS/etc/gui/wxpython/wxgui.py", line 87, in <module>
   import gui_modules.nviz_tools as nviz_tools
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/nviz_tools.py", line 36, in <module>
   from nviz_mapdisp import wxUpdateView as wxUpdateView
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/nviz_mapdisp.py", line 41, in <module>
   import wxnviz
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/wxnviz.py", line 24, in <module>
   from grass.lib.grass import *
ImportError: No module named lib.grass

tring to re-run : "g.gui" i have the same log.

just to check if my python "is bugged"
i build python2.6.5 source code,
in a non standard location "/usr/local/gislib/unix",
as unix library (no framework)
then i exported the path for the new python
and i used it to reconfigure-compile-install grass70,
but unluky i have problems to build wxpython
(i compiled wxwidjet as 32+64 bit but i'm having
problems to built wxpython on it)
i'll try to figure out a wxpython cocoa build
and tell you if the grass.lib not found problem persists.

thanks for any help.

Il giorno 27/mag/2010, alle ore 16.52, Michael Barton ha scritto:

No ipython on the student's machine.

I compiled GRASS 7 and had no problems. I posted my binary so others could use it. The one person who has tried it gets a bogus error about Python subprocess() not existing. The binary of GRASS 6.5 that I compiled and posted at the same time runs fine on her machine (demonstrating that subprocess() is OK).

Oddly, some other folks here got similar bogus errors with GRASS 6.5 or 7 (I can't remember) when trying to run a Python script under winGRASS a few weeks back.

Michael
____________________
C. Michael Barton
Director, Center for Social Dynamics & Complexity
Professor of Anthropology, School of Human Evolution & Social Change
Arizona State University

voice: 480-965-6262 (SHESC), 480-727-9746 (CSDC)
fax: 480-965-7671 (SHESC), 480-727-0709 (CSDC)
www: www.public.asu.edu/~cmbarton, http://csdc.asu.edu

On May 27, 2010, at 6:43 AM, William Kyngesburye wrote:

On May 27, 2010, at 3:56 AM, Massimo Di Stefano <massimodisasha@gmail.com

wrote:

Hell Soeren,

seems i've an up-to-date revision :

MacBook-Pro-15-di-Massimo-Di-Stefano:grass_trunk sasha$ svn up
At revision 42358.

i also did make distclean first to re-try configure && make.

but your note about a "renamed" function, point me to a problem i
had on qgis side too (make install overwrite the .app but ... it
don't do it as aspected)

infact, removing manually the old grass70.app (compiled from source)
trashing the complete app from /Application , instead of overwrite
it using "sudo make install"

now i have :

In [1]: import os, sys

In [2]: from grass.lib import raster
------------------------------------------------------------
Traceback (most recent call last):
File "<ipython console>", line 1, in <module>
ImportError: No module named lib

I wonder if ipython is affecting this? Just a random idea. (I can't
try anything right now- dev Mac died)

the error is present but changes ... this time maybe can be a wrong
sintax i used ???
weird .. i'm a bit confused :wink:

i'm tring to replicate, from command line, the code :

###

import os, sys
from grass.lib import grass, raster
from ctypes import *

input = sys.argv[1]

grass.G_gisinit(sys.argv[0])

mapset = grass.G_find_raster2(input, "")
print mapset

Massimo

Il giorno 27/mag/2010, alle ore 10.37, Soeren Gebbert ha scritto:

Hello Massimo,
can you please make a svn update?The gmath function
G_math_backward_solving has been renamed into
G_math_backward_substitution a while ago. Maybe something got mixed
up?

Best
Soeren

12

Hi,

2010/6/1 Massimo Di Stefano <massimodisasha@gmail.com>:

[...]

Traceback (most recent call last):
File "/Applications/GRASS-7.0.app/Contents/MacOS/etc/gui/wxpython/wxgui.py", line 87, in <module>
import gui_modules.nviz_tools as nviz_tools
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/nviz_tools.py", line 36, in <module>
from nviz_mapdisp import wxUpdateView as wxUpdateView
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/nviz_mapdisp.py", line 41, in <module>
import wxnviz
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/wxnviz.py", line 24, in <module>
from grass.lib.grass import *
ImportError: No module named lib.grass

ctypes grass.lib (lib/python/ctypes) is missing in your installation
(ctypesgen required). wxGUI failure fixed in r42430.

Martin

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

13

On Jun 1, 2010, at 5:52 PM, Barton Michael wrote:

I just compiled GRASS 7 a couple days ago, before leaving for Spain. Wx digitizer and nviz don't compile.

I had the same problem two days ago - I will update and try again.
Everything else runs so far but I really wanted to check on the new wxnviz development.

Helena

Also others can't even open my binaries.

Is there a new configure switch for ctypes? Are we supposed to no longer use the python switch?

Michael
____________________
C. Michael Barton
Director, Center for Social Dynamics & Complexity
Professor of Anthropology, School of Human Evolution & Social Change
Arizona State University

voice: 480-965-6262 (SHESC), 480-727-9746 (CSDC)
fax: 480-965-7671 (SHESC), 480-727-0709 (CSDC)
www: www.public.asu.edu/~cmbarton, http://csdc.asu.edu

On Jun 1, 2010, at 11:36 PM, Massimo Di Stefano wrote:

Now the error i means :

"""

In [2]: from grass.lib import raster
------------------------------------------------------------
Traceback (most recent call last):
File "<ipython console>", line 1, in <module>
ImportError: No module named lib

"""

is reflected in the gui.
the new GUI code changes are affected by it,
tring to build grass70, starting the application
i have a broken gui that show me this error log :

GRASS 7.0.svn (spearfish60):~ >
WARNING: Vector digitizer is not available (dlopen(/Applications/GRASS-7.0.app/Contents/MacOS/etc/gui/wxpython/vdigit/_grass7_wxvdigit.so, 2): Symbol not found: __Py_RefTotal
Referenced from: /Applications/GRASS-7.0.app/Contents/MacOS/etc/gui/wxpython/vdigit/_grass7_wxvdigit.so
Expected in: flat namespace
in /Applications/GRASS-7.0.app/Contents/MacOS/etc/gui/wxpython/vdigit/_grass7_wxvdigit.so).

Note that the vector digitizer is currently not working under MS Windows (hopefully this will be fixed soon). Please keep an eye out for updated versions of GRASS.
Traceback (most recent call last):
File "/Applications/GRASS-7.0.app/Contents/MacOS/etc/gui/wxpython/wxgui.py", line 87, in <module>
  import gui_modules.nviz_tools as nviz_tools
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/nviz_tools.py", line 36, in <module>
  from nviz_mapdisp import wxUpdateView as wxUpdateView
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/nviz_mapdisp.py", line 41, in <module>
  import wxnviz
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/wxnviz.py", line 24, in <module>
  from grass.lib.grass import *
ImportError: No module named lib.grass

tring to re-run : "g.gui" i have the same log.

just to check if my python "is bugged"
i build python2.6.5 source code,
in a non standard location "/usr/local/gislib/unix",
as unix library (no framework)
then i exported the path for the new python
and i used it to reconfigure-compile-install grass70,
but unluky i have problems to build wxpython
(i compiled wxwidjet as 32+64 bit but i'm having
problems to built wxpython on it)
i'll try to figure out a wxpython cocoa build
and tell you if the grass.lib not found problem persists.

thanks for any help.

Il giorno 27/mag/2010, alle ore 16.52, Michael Barton ha scritto:

No ipython on the student's machine.

I compiled GRASS 7 and had no problems. I posted my binary so others could use it. The one person who has tried it gets a bogus error about Python subprocess() not existing. The binary of GRASS 6.5 that I compiled and posted at the same time runs fine on her machine (demonstrating that subprocess() is OK).

Oddly, some other folks here got similar bogus errors with GRASS 6.5 or 7 (I can't remember) when trying to run a Python script under winGRASS a few weeks back.

Michael
____________________
C. Michael Barton
Director, Center for Social Dynamics & Complexity
Professor of Anthropology, School of Human Evolution & Social Change
Arizona State University

voice: 480-965-6262 (SHESC), 480-727-9746 (CSDC)
fax: 480-965-7671 (SHESC), 480-727-0709 (CSDC)
www: www.public.asu.edu/~cmbarton, http://csdc.asu.edu

On May 27, 2010, at 6:43 AM, William Kyngesburye wrote:

On May 27, 2010, at 3:56 AM, Massimo Di Stefano <massimodisasha@gmail.com

wrote:

Hell Soeren,

seems i've an up-to-date revision :

MacBook-Pro-15-di-Massimo-Di-Stefano:grass_trunk sasha$ svn up
At revision 42358.

i also did make distclean first to re-try configure && make.

but your note about a "renamed" function, point me to a problem i
had on qgis side too (make install overwrite the .app but ... it
don't do it as aspected)

infact, removing manually the old grass70.app (compiled from source)
trashing the complete app from /Application , instead of overwrite
it using "sudo make install"

now i have :

In [1]: import os, sys

In [2]: from grass.lib import raster
------------------------------------------------------------
Traceback (most recent call last):
File "<ipython console>", line 1, in <module>
ImportError: No module named lib

I wonder if ipython is affecting this? Just a random idea. (I can't
try anything right now- dev Mac died)

the error is present but changes ... this time maybe can be a wrong
sintax i used ???
weird .. i'm a bit confused :wink:

i'm tring to replicate, from command line, the code :

###

import os, sys
from grass.lib import grass, raster
from ctypes import *

input = sys.argv[1]

grass.G_gisinit(sys.argv[0])

mapset = grass.G_find_raster2(input, "")
print mapset

Massimo

Il giorno 27/mag/2010, alle ore 10.37, Soeren Gebbert ha scritto:

Hello Massimo,
can you please make a svn update?The gmath function
G_math_backward_solving has been renamed into
G_math_backward_substitution a while ago. Maybe something got mixed
up?

Best
Soeren

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

14

I succesfull build wxpython on snow leopard to be 32+64 bit against wx2.9 (svn version) <- is the only way to have wx with no "bitwise" problems on osx
then i installed ctypesgen

tring to build grass70 i got the error in wxpython gui code
re-running make inside its directory the error is not fixed.

i suppose this is related to wx version (2.9) instead of (2.8)
so it shouldn't be a grass related problem ...

this is the log :

http://www.geofemengineering.it/data/grass_wxgui_log.txt

i wonder if it's possible to adapt grass-gui code to run under 2.9 too :-/

i'm now recompiling ctypsegen on the system python to sede if it fix the grass.lib error

Il giorno 01/giu/2010, alle ore 23.53, Martin Landa ha scritto:

Hi,

2010/6/1 Massimo Di Stefano <massimodisasha@gmail.com>:

[...]

Traceback (most recent call last):
File "/Applications/GRASS-7.0.app/Contents/MacOS/etc/gui/wxpython/wxgui.py", line 87, in <module>
   import gui_modules.nviz_tools as nviz_tools
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/nviz_tools.py", line 36, in <module>
   from nviz_mapdisp import wxUpdateView as wxUpdateView
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/nviz_mapdisp.py", line 41, in <module>
   import wxnviz
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/wxnviz.py", line 24, in <module>
   from grass.lib.grass import *
ImportError: No module named lib.grass

ctypes grass.lib (lib/python/ctypes) is missing in your installation
(ctypesgen required). wxGUI failure fixed in r42430.

Martin

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

15

Barton Michael wrote:

I just compiled GRASS 7 a couple days ago, before leaving for Spain. Wx
digitizer and nviz don't compile.

The gui/wxpython/nviz directory is no longer built; it has been
superseded by gui/wxpython/gui_modules/wxnviz.py.

The ctypes/SWIG changes don't touch the vdigit module.

Also others can't even open my
binaries.

Is there a new configure switch for ctypes?

No. I added a local copy of ctypesgen to the source tree, so there are
no external dependencies.

ctypes itself is part of the standard library in Python 2.5 and 2.6.

Are we supposed to no longer
use the python switch?

--with-python should only be required for vdigit.

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

16

Il giorno 02/giu/2010, alle ore 04.54, Glynn Clements ha scritto:

Barton Michael wrote:

I just compiled GRASS 7 a couple days ago, before leaving for Spain. Wx
digitizer and nviz don't compile.

The gui/wxpython/nviz directory is no longer built; it has been
superseded by gui/wxpython/gui_modules/wxnviz.py.

The ctypes/SWIG changes don't touch the vdigit module.

Also others can't even open my
binaries.

Is there a new configure switch for ctypes?

No. I added a local copy of ctypesgen to the source tree, so there are
no external dependencies.

ctypes itself is part of the standard library in Python 2.5 and 2.6.

This means that also on OSX we don't need to install external libraries beacouse WxPython, Ctype are included in the system Python
and Ctypsgen is included in the Grass source (hence i don't have to install it as suggested in the preciouse mails)
At this point i'm without any clue on how to debug the grass7.0 problems on OSX.
I tried installing a new version of Python but unlucky it fails to build the gui (maybe beacouse i'm using wx2.9, see my previouse mail)

i tried to compile grass using python.framework this time,
but seems it find only python2.61 (system python)
i treid to change grass.sh script .. but without lucky
so i chosed to build without .app generation
the results is :

error in wx gui build (i'm using wx-cocoa)
log :
http://www.geofemengineering.it/data/grass_wxgui_log.txt

while start grass in text mode, again grass.lib is not avaiable as grass.python module
(i have tried with (installed manually from svn) and without ctypesgen (using the version included in grass))

Have you any suggestion like
"add a print statment in the source code file X at line Y "
to try to see where the problem begins ?

Are we supposed to no longer
use the python switch?

--with-python should only be required for vdigit.

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

17

Massimo Di Stefano wrote:

error in wx gui build (i'm using wx-cocoa)
log :
http://www.geofemengineering.it/data/grass_wxgui_log.txt

pseudodc.h: In member function ‘virtual void gpdcFloodFillOp::DrawToDC(wxDC*, bool)’:
pseudodc.h:210: error: invalid conversion from ‘int’ to ‘wxFloodFillStyle’
pseudodc.h:210: error: initializing argument 4 of ‘bool wxDC::FloodFill(wxCoord, wxCoord, const wxColour&, wxFloodFillStyle)’

These are caused by trying to use wxWidgets 2.9. GRASS requires
wxWidgets 2.8.

while start grass in text mode, again grass.lib is not avaiable as
grass.python module

You'll need to provide the output from "make -C lib/python" if you
want help with this.

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

18

Il giorno 02/giu/2010, alle ore 10.41, Glynn Clements ha scritto:

Massimo Di Stefano wrote:

error in wx gui build (i'm using wx-cocoa)
log :
http://www.geofemengineering.it/data/grass_wxgui_log.txt

pseudodc.h: In member function ‘virtual void gpdcFloodFillOp::DrawToDC(wxDC*, bool)’:
pseudodc.h:210: error: invalid conversion from ‘int’ to ‘wxFloodFillStyle’
pseudodc.h:210: error: initializing argument 4 of ‘bool wxDC::FloodFill(wxCoord, wxCoord, const wxColour&, wxFloodFillStyle)’

These are caused by trying to use wxWidgets 2.9. GRASS requires
wxWidgets 2.8.

yes i suspected the error was related to wx version
mine was just an attempt, beacouse i had to build it
for the new python

while start grass in text mode, again grass.lib is not avaiable as
grass.python module

You'll need to provide the output from "make -C lib/python" if you
want help with this.

Thanks! i see this log thinking about the wxgui failure ... while was grass lib,
this the log :

http://www.geofemengineering.it/data/grass_lib_python_osx_log.txt

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

19

It is not a wxPython 2.9 problem--or should not be. I think it is something else in the build system. Others are having the same problem with the current version.

Michael
____________________
C. Michael Barton
Director, Center for Social Dynamics & Complexity
Professor of Anthropology, School of Human Evolution & Social Change
Arizona State University

voice: 480-965-6262 (SHESC), 480-727-9746 (CSDC)
fax: 480-965-7671 (SHESC), 480-727-0709 (CSDC)
www: www.public.asu.edu/~cmbarton, http://csdc.asu.edu

On Jun 2, 2010, at 2:34 AM, Massimo Di Stefano wrote:

I succesfull build wxpython on snow leopard to be 32+64 bit against wx2.9 (svn version) <- is the only way to have wx with no "bitwise" problems on osx
then i installed ctypesgen

tring to build grass70 i got the error in wxpython gui code
re-running make inside its directory the error is not fixed.

i suppose this is related to wx version (2.9) instead of (2.8)
so it shouldn't be a grass related problem ...

this is the log :

http://www.geofemengineering.it/data/grass_wxgui_log.txt

i wonder if it's possible to adapt grass-gui code to run under 2.9 too :-/

i'm now recompiling ctypsegen on the system python to sede if it fix the grass.lib error

Il giorno 01/giu/2010, alle ore 23.53, Martin Landa ha scritto:

Hi,

2010/6/1 Massimo Di Stefano <massimodisasha@gmail.com>:

[...]

Traceback (most recent call last):
File "/Applications/GRASS-7.0.app/Contents/MacOS/etc/gui/wxpython/wxgui.py", line 87, in <module>
  import gui_modules.nviz_tools as nviz_tools
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/nviz_tools.py", line 36, in <module>
  from nviz_mapdisp import wxUpdateView as wxUpdateView
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/nviz_mapdisp.py", line 41, in <module>
  import wxnviz
File "/Users/Shared/source/grass_trunk/dist.i386-apple-darwin10.3.0/etc/gui/wxpython/gui_modules/wxnviz.py", line 24, in <module>
  from grass.lib.grass import *
ImportError: No module named lib.grass

ctypes grass.lib (lib/python/ctypes) is missing in your installation
(ctypesgen required). wxGUI failure fixed in r42430.

Martin

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

20

When the switch to ctypes started, I compiled ctypes as directed. Could this be a problem now?

Michael
____________________
C. Michael Barton
Director, Center for Social Dynamics & Complexity
Professor of Anthropology, School of Human Evolution & Social Change
Arizona State University

voice: 480-965-6262 (SHESC), 480-727-9746 (CSDC)
fax: 480-965-7671 (SHESC), 480-727-0709 (CSDC)
www: www.public.asu.edu/~cmbarton, http://csdc.asu.edu

On Jun 2, 2010, at 4:54 AM, Glynn Clements wrote:

Barton Michael wrote:

I just compiled GRASS 7 a couple days ago, before leaving for Spain. Wx
digitizer and nviz don't compile.

The gui/wxpython/nviz directory is no longer built; it has been
superseded by gui/wxpython/gui_modules/wxnviz.py.

The ctypes/SWIG changes don't touch the vdigit module.

Also others can't even open my
binaries.

Is there a new configure switch for ctypes?

No. I added a local copy of ctypesgen to the source tree, so there are
no external dependencies.

ctypes itself is part of the standard library in Python 2.5 and 2.6.

Are we supposed to no longer
use the python switch?

--with-python should only be required for vdigit.

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