Menu
▾
▴
visualpython-users
|
From:
<fre...@gb...> - 2007-12-21 23:02:02
|
Hi, I'm using vpython under debian lenny, and as soon as I close the window, or= =20 call the display hide() method, my app crash with the following error: Gdk-ERROR **: BadValue (integer parameter out of range for operation) serial 1309 error_code 2 request_code 129 minor_code 9 Am I doing something wrong? ASFAIR, I had this problem long ago, when I=20 tried vptyhon for the first time... Thanks, =2D-=20 Fr=E9d=E9ric http://www.gbiloba.org |
|
From:
<fre...@gb...> - 2007-12-21 23:39:31
|
On samedi 22 d=E9cembre 2007, Fr=E9d=E9ric Mantegazza wrote: > I'm using vpython under debian lenny, and as soon as I close the window, > or call the display hide() method, my app crash with the following > error: > > Gdk-ERROR **: BadValue (integer parameter out of range for operation) > serial 1309 error_code 2 request_code 129 minor_code 9 > > Am I doing something wrong? ASFAIR, I had this problem long ago, when I > tried vptyhon for the first time... I just checked on a Windows XP running under VirtualBox: no crash there.=20 This is a linux-only problem... =2D-=20 Fr=E9d=E9ric http://www.gbiloba.org |
|
From: Bruce S. <Bru...@nc...> - 2007-12-22 06:05:23
|
This isn't a familiar-sounding problem. What version of VPython is this? Did you build it from a tarball or obtain it from an existing Debian package? When you say "the display hide() method", do you mean scene.visible = False, or something else? VPython is known to work on a variety of Linux platforms, so it isn't something as simple as "it doesn't work on Linux". Bruce Sherwood Frédéric Mantegazza wrote: > On samedi 22 décembre 2007, Frédéric Mantegazza wrote: > > >> I'm using vpython under debian lenny, and as soon as I close the window, >> or call the display hide() method, my app crash with the following >> error: >> >> Gdk-ERROR **: BadValue (integer parameter out of range for operation) >> serial 1309 error_code 2 request_code 129 minor_code 9 >> >> Am I doing something wrong? ASFAIR, I had this problem long ago, when I >> tried vptyhon for the first time... >> > > I just checked on a Windows XP running under VirtualBox: no crash there. > This is a linux-only problem... > > |
|
From:
<fre...@gb...> - 2007-12-22 08:25:20
|
On samedi 22 d=E9cembre 2007, Bruce Sherwood wrote:
> This isn't a familiar-sounding problem. What version of VPython is this?
> Did you build it from a tarball or obtain it from an existing Debian
> package? When you say "the display hide() method", do you mean
> scene.visible =3D False, or something else?
Sorry, I forgot all that usefull informations. I'm using vpython 3.2.9,=20
compiled from source (I would like to try version 4, but some lib are=20
missing under debian; I'll have to get them from sources).
The crash occurs as soon as I click on the window close box, or if I call=20
scene.hide() or use scene.visible =3D False, but only if there is at least=
=20
one object in the scene; an empty scene does not crash the program.
I may make a mistake using vpython, but the same code runs fine under=20
Windows.
> VPython is known to work on a variety of Linux platforms, so it isn't
> something as simple as "it doesn't work on Linux".
As I said, I'm using debian (lenny). And I'm sure I had this problem some=20
years ago, when I first used vpython (also under debian, but stable sarge)=
=20
for my work: a student made a great 3D view of a complete neutron=20
spectrometer, with real time refresh:
http://www.gbiloba.org/download/SNP-definitif-640x480.avi
http://www.gbiloba.org/download/LPA-definitif-640x480.avi
He used it for its presentation, and people where amazed ;o)
Ok. I attached all build logs, and a simple script which fails. I ran it=20
with strace -f (to see all threads).
Thanks for you help.
=2D-=20
Fr=E9d=E9ric
http://www.gbiloba.org
|
|
From:
<fre...@gb...> - 2007-12-22 10:55:59
|
On samedi 22 d=E9cembre 2007, Bruce Sherwood wrote: > This isn't a familiar-sounding problem. What version of VPython is this? > Did you build it from a tarball or obtain it from an existing Debian > package? When you say "the display hide() method", do you mean > scene.visible =3D False, or something else? > > VPython is known to work on a variety of Linux platforms, so it isn't > something as simple as "it doesn't work on Linux". Ok, I made some more tests. My wife as the same computer, same debian lenny= =20 up-to-date, but a different graphics card. Mine is a ATI Radeon M9 (r200)=20 (working with free radeon driver), hers is a NVidia, with NVidia drivers. Here are the results: =2D vpython works fine on her computer =2D vpython works fine if I connect to my computer from hers (using ssh -X) =2D vpython works fine if I connect to her computer from mine (using ssh -X) =2D vpython crashes if I launch it from my computer So, there is something related to my video card driver, but I don't know=20 what. I would have understood if vpython always crashed when I *display*=20 it on *my* computer (directly running it on my computer, or running it on=20 my wife's one, throught ssh -X), as OpenGL stuff used is the one where the= =20 display occurs... But here, it is a little different. Does someone has an idea? How can I make further tests to find the problem? Thanks, =2D-=20 Fr=E9d=E9ric http://www.gbiloba.org |
|
From: Bruce S. <Bru...@nc...> - 2007-12-22 17:55:51
|
Definitely an interesting situation. Presumably there's something wrong with your graphics card driver, as you say, and the only fix is probably an updated driver (which of course may not exist). Perhaps VPython runs as a remote display because the timing is somewhat different. Concerning trying the beta version, what libraries are you missing? Bruce Sherwood Frédéric Mantegazza wrote: > On samedi 22 décembre 2007, Bruce Sherwood wrote: > > >> This isn't a familiar-sounding problem. What version of VPython is this? >> Did you build it from a tarball or obtain it from an existing Debian >> package? When you say "the display hide() method", do you mean >> scene.visible = False, or something else? >> >> VPython is known to work on a variety of Linux platforms, so it isn't >> something as simple as "it doesn't work on Linux". >> > > Ok, I made some more tests. My wife as the same computer, same debian lenny > up-to-date, but a different graphics card. Mine is a ATI Radeon M9 (r200) > (working with free radeon driver), hers is a NVidia, with NVidia drivers. > > Here are the results: > > - vpython works fine on her computer > - vpython works fine if I connect to my computer from hers (using ssh -X) > - vpython works fine if I connect to her computer from mine (using ssh -X) > - vpython crashes if I launch it from my computer > > So, there is something related to my video card driver, but I don't know > what. I would have understood if vpython always crashed when I *display* > it on *my* computer (directly running it on my computer, or running it on > my wife's one, throught ssh -X), as OpenGL stuff used is the one where the > display occurs... But here, it is a little different. > > Does someone has an idea? How can I make further tests to find the problem? > > Thanks, > > |
|
From:
<fre...@gb...> - 2007-12-22 20:39:25
|
On samedi 22 d=E9cembre 2007, Bruce Sherwood wrote: > Definitely an interesting situation. Presumably there's something wrong > with your graphics card driver, as you say, and the only fix is probably > an updated driver (which of course may not exist). Well, I'm always up-to-date with debian testing... > Perhaps VPython runs as a remote display because the timing is somewhat=20 > different.=20 No idea how I can dig the problem? Where can I add debug outputs in your=20 code to find what happens?=20 > Concerning trying the beta version, what libraries are you missing? configure: error: gtkglextmm 1.2, pangoft2, glibmm-2.4, and pangomm-1.4=20 libglademm-2.4 are required on Unix-like systems =2D-=20 Fr=E9d=E9ric http://www.gbiloba.org |
|
From: Bruce S. <Bru...@nc...> - 2007-12-23 00:44:40
|
Here is a link to updated instructions for installing the beta version: http://vpython.org/INSTALL.html This may be a bit more up to date than the INSTALL.txt file included in the tarball. It tries to do a better job of listing dependencies and how to deal with them. I don't think there's any debugging code you can put in the Visual source code that would help you with what is almost certainly a graphics driver problem. I think some of your missing libraries are included in some of those listed in the INSTALL file, but if you find that not to be the case I'd appreciate knowing so that the file can be improved. Bruce Sherwood Frédéric Mantegazza wrote: > On samedi 22 décembre 2007, Bruce Sherwood wrote: > > >> Definitely an interesting situation. Presumably there's something wrong >> with your graphics card driver, as you say, and the only fix is probably >> an updated driver (which of course may not exist). >> > > Well, I'm always up-to-date with debian testing... > > >> Perhaps VPython runs as a remote display because the timing is somewhat >> different. >> > > No idea how I can dig the problem? Where can I add debug outputs in your > code to find what happens? > > >> Concerning trying the beta version, what libraries are you missing? >> > > configure: error: gtkglextmm 1.2, pangoft2, glibmm-2.4, and pangomm-1.4 > libglademm-2.4 are required on Unix-like systems > > |
|
From:
<fre...@gb...> - 2007-12-23 11:09:37
|
On dimanche 23 d=E9cembre 2007, Bruce Sherwood wrote:
> I think some of your missing libraries are included in some of those
> listed in the INSTALL file, but if you find that not to be the case I'd
> appreciate knowing so that the file can be improved.
In fact, the only package missing was gtkglextmm; all others where=20
correctly installed. The problem comes from the configure script. When it=20
call pkg-config, to check mising packages, it gets:
configure:19706: checking for pkg-config
configure:19724: found /usr/bin/pkg-config
configure:19736: result: /usr/bin/pkg-config
configure:19765: checking pkg-config is at least version 0.9.0
configure:19768: result: yes
configure:19792: checking for GTK
configure:19800: $PKG_CONFIG --exists --print-errors "gtkglextmm-1.2 >=3D 1=
=2E2
pangoft2 glibmm-2.4 pangomm-1.4 libglademm-2.4 freetype2"
Package gtkglextmm-1.2 was not found in the pkg-config search path.
Perhaps you should add the directory containing `gtkglextmm-1.2.pc'
to the PKG_CONFIG_PATH environment variable
No package 'gtkglextmm-1.2' found
configure:19803: $? =3D 1
configure:19818: $PKG_CONFIG --exists --print-errors "gtkglextmm-1.2 >=3D 1=
=2E2=20
pangoft2 glibmm-2.4 pangomm-1.4 libglademm-2.4 freetype2"
Package gtkglextmm-1.2 was not found in the pkg-config search path.
Perhaps you should add the directory containing `gtkglextmm-1.2.pc'
to the PKG_CONFIG_PATH environment variable
No package 'gtkglextmm-1.2' found
configure:19821: $? =3D 1
No package 'gtkglextmm-1.2' found
configure:19849: result: no
configure:19851: error: gtkglextmm 1.2, pangoft2, glibmm-2.4, and=20
pangomm-1.4 libglademm-2.4 are required on Unix-like systems
As you can see, it outputs all libs, even if the only missing is=20
gtkglextmm. I compiled and installed it, and it now works fine :o)
But during the compilation phase, I still had problems of missing libs:=20
libboost-python (and -thread) is not checked in the configure. Once=20
installed, it works (it was the same in previous vpython versions).
And I have a another problem: vpython can't find the numpy/arrayobject.h=20
(imported from include/python/num_util.hpp). This file is in:
python-numpy:
usr/lib/python2.4/site-packages/numpy/core/include/numpy/arrayobject.h
But the configure script found an incorrect path:
=20
g++ -I/usr/include/python2.4 -I/usr/local/lib/python2.4/site-packages/numpy=
/core/include -DHAVE_CONFIG_H -I../include -I../include -I/usr/local/includ
e/gtkglextmm-1.2 -I/usr/local/lib/gtkglextmm-1.2/include -I/usr/include/gtk=
glext-1.0 -I/usr/include/gtkmm-2.4 -I/usr/lib/gtkmm-2.4/include -I/usr/lib/g
tkglext-1.0/include -I/usr/include/gdkmm-2.4 -I/usr/lib/gdkmm-2.4/include -=
I/usr/include/pangomm-1.4 -I/usr/include/gtk-2.0 -I/usr/lib/gtk-2.0/include
=2DI/usr/include/pango-1.0 -I/usr/include/glib-2.0 -I/usr/lib/glib-2.0/incl=
ude -I/usr/include/cairo -I/usr/include/freetype2 -I/usr/include/libpng12 -=
I/u
sr/include/glibmm-2.4 -I/usr/lib/glibmm-2.4/include -I/usr/include/cairomm-=
1.0 -I/usr/include/sigc++-2.0 -I/usr/lib/sigc++-2.0/include -I/usr/include/a
tk-1.0 -I/usr/include/atkmm-1.6 -I/usr/include/libglademm-2.4 -I/usr/lib/li=
bglademm-2.4/include -I/usr/include/libglade-2.0 -I/usr/include/libxml2 -I..
/include/gtk2 -pthread -I/usr/include/glib-2.0 -I/usr/lib/glib-2.0/include =
=2DI/usr/include/python2.4 -I/usr/local/lib/python2.4/site-packages/numpy/c=
ore
/include -fpic -DPIC -g -O2 -ftemplate-depth-120 -MMD -MF=20
convex.d -MT "convex.d=20
convex.lo" -c ./python/convex.cpp -fPIC -DPIC -o .libs/convex.o
In file included from ../include/python/convex.hpp:12,
from ./python/convex.cpp:6:
=2E./include/python/num_util.hpp:68:31: warning: numpy/arrayobject.h: No su=
ch=20
file or directory
It gave /usr/local/lib/..., instead of /usr/lib/... I don't have any numpy=
=20
package in /usr/local/lib... I made a symbolic link, and it compiled, but=20
it is not very clean.
=2D-=20
Fr=E9d=E9ric
http://www.gbiloba.org
|
|
From: Bruce S. <Bru...@nc...> - 2007-12-25 06:19:38
|
Thanks for the detailed report. I don't expect to be able very soon to address the issue of better reporting of precisely which libraries are missing, but some day.... I'm puzzled about the boost libraries, as I thought those were checked. The problem with numpy sounds like an error in issuing the configure command: you need configure --prefix=/usr if Python and its site-packages are in /usr. As is documented in the install instructions, by default configure assumes the target Python is in /usr/local. Bruce Sherwood Frédéric Mantegazza wrote: > On dimanche 23 décembre 2007, Bruce Sherwood wrote: > > >> I think some of your missing libraries are included in some of those >> listed in the INSTALL file, but if you find that not to be the case I'd >> appreciate knowing so that the file can be improved. >> > > In fact, the only package missing was gtkglextmm; all others where > correctly installed. The problem comes from the configure script. When it > call pkg-config, to check mising packages, it gets: > > configure:19706: checking for pkg-config > configure:19724: found /usr/bin/pkg-config > configure:19736: result: /usr/bin/pkg-config > configure:19765: checking pkg-config is at least version 0.9.0 > configure:19768: result: yes > configure:19792: checking for GTK > configure:19800: $PKG_CONFIG --exists --print-errors "gtkglextmm-1.2 >= 1.2 > pangoft2 glibmm-2.4 pangomm-1.4 libglademm-2.4 freetype2" > Package gtkglextmm-1.2 was not found in the pkg-config search path. > Perhaps you should add the directory containing `gtkglextmm-1.2.pc' > to the PKG_CONFIG_PATH environment variable > No package 'gtkglextmm-1.2' found > configure:19803: $? = 1 > configure:19818: $PKG_CONFIG --exists --print-errors "gtkglextmm-1.2 >= 1.2 > pangoft2 glibmm-2.4 pangomm-1.4 libglademm-2.4 freetype2" > Package gtkglextmm-1.2 was not found in the pkg-config search path. > Perhaps you should add the directory containing `gtkglextmm-1.2.pc' > to the PKG_CONFIG_PATH environment variable > No package 'gtkglextmm-1.2' found > configure:19821: $? = 1 > No package 'gtkglextmm-1.2' found > configure:19849: result: no > configure:19851: error: gtkglextmm 1.2, pangoft2, glibmm-2.4, and > pangomm-1.4 libglademm-2.4 are required on Unix-like systems > > As you can see, it outputs all libs, even if the only missing is > gtkglextmm. I compiled and installed it, and it now works fine :o) > > But during the compilation phase, I still had problems of missing libs: > libboost-python (and -thread) is not checked in the configure. Once > installed, it works (it was the same in previous vpython versions). > > And I have a another problem: vpython can't find the numpy/arrayobject.h > (imported from include/python/num_util.hpp). This file is in: > > python-numpy: > usr/lib/python2.4/site-packages/numpy/core/include/numpy/arrayobject.h > > But the configure script found an incorrect path: > > g++ -I/usr/include/python2.4 -I/usr/local/lib/python2.4/site-packages/numpy/core/include -DHAVE_CONFIG_H -I../include -I../include -I/usr/local/includ > e/gtkglextmm-1.2 -I/usr/local/lib/gtkglextmm-1.2/include -I/usr/include/gtkglext-1.0 -I/usr/include/gtkmm-2.4 -I/usr/lib/gtkmm-2.4/include -I/usr/lib/g > tkglext-1.0/include -I/usr/include/gdkmm-2.4 -I/usr/lib/gdkmm-2.4/include -I/usr/include/pangomm-1.4 -I/usr/include/gtk-2.0 -I/usr/lib/gtk-2.0/include > -I/usr/include/pango-1.0 -I/usr/include/glib-2.0 -I/usr/lib/glib-2.0/include -I/usr/include/cairo -I/usr/include/freetype2 -I/usr/include/libpng12 -I/u > sr/include/glibmm-2.4 -I/usr/lib/glibmm-2.4/include -I/usr/include/cairomm-1.0 -I/usr/include/sigc++-2.0 -I/usr/lib/sigc++-2.0/include -I/usr/include/a > tk-1.0 -I/usr/include/atkmm-1.6 -I/usr/include/libglademm-2.4 -I/usr/lib/libglademm-2.4/include -I/usr/include/libglade-2.0 -I/usr/include/libxml2 -I.. > /include/gtk2 -pthread -I/usr/include/glib-2.0 -I/usr/lib/glib-2.0/include -I/usr/include/python2.4 -I/usr/local/lib/python2.4/site-packages/numpy/core > /include -fpic -DPIC -g -O2 -ftemplate-depth-120 -MMD -MF > convex.d -MT "convex.d > convex.lo" -c ./python/convex.cpp -fPIC -DPIC -o .libs/convex.o > In file included from ../include/python/convex.hpp:12, > from ./python/convex.cpp:6: > ../include/python/num_util.hpp:68:31: warning: numpy/arrayobject.h: No such > file or directory > > It gave /usr/local/lib/..., instead of /usr/lib/... I don't have any numpy > package in /usr/local/lib... I made a symbolic link, and it compiled, but > it is not very clean. > > |
|
From:
<fre...@gb...> - 2007-12-25 10:14:05
|
On mardi 25 d=E9cembre 2007, Bruce Sherwood wrote: > Thanks for the detailed report. I don't expect to be able very soon to > address the issue of better reporting of precisely which libraries are > missing, but some day.... I'm puzzled about the boost libraries, as I > thought those were checked. Ok. > The problem with numpy sounds like an error=20 > in issuing the configure command: you need configure --prefix=3D/usr if > Python and its site-packages are in /usr. As is documented in the > install instructions, by default configure assumes the target Python is > in /usr/local. I understand, but I want to install manually compiled packages=20 in /usr/local, so I set --prefix to that dir. I may have to give another=20 param to tell configure to search dependent libs? Anyway, it is not a big=20 problem. =2D-=20 Fr=E9d=E9ric http://www.gbiloba.org |
|
From: Bruce S. <Bru...@nc...> - 2007-12-25 16:17:02
|
There are two different issues: where is Python, and where will visual be installed. If you said --prefix=/usr/local but didn't specify that Python was to be found in /usr, you would have a problem. From the Linux download page at vpython.org: *which python* (find out where Python is located)* * Make a note of the /prefix/ preceding /bin/python, such as /usr or /usr/local. (a) If /prefix/ is /usr/local, execute *../visual-x.x.x/configure* (b) If /prefix/ is something else, and Visual can go into prefix/lib/python/site-packages, execute * ../visual-x.x.x/configure --prefix=*/prefix/ (c) If you want to use a different version of Python than the one found with "which python", or (b) is not appropriate, specify both the particular Python and where to install Visual: * PYTHON=/somewhere1/bin/python ../visual-x.x.x/configure --prefix=/somewhere2* If "somewhere1" and "somewhere2" are different, you must also add the "somewhere2" directory to Python's module search path. For details, at www.python.org <http://www.python.org> read section 4.1 (Modifying Python's Search Path) in the section Installing Python Modules of the Python 2.3 on-line documentation. Bruce Sherwood Frédéric Mantegazza wrote: > On mardi 25 décembre 2007, Bruce Sherwood wrote: > > >> Thanks for the detailed report. I don't expect to be able very soon to >> address the issue of better reporting of precisely which libraries are >> missing, but some day.... I'm puzzled about the boost libraries, as I >> thought those were checked. >> > > Ok. > > >> The problem with numpy sounds like an error >> in issuing the configure command: you need configure --prefix=/usr if >> Python and its site-packages are in /usr. As is documented in the >> install instructions, by default configure assumes the target Python is >> in /usr/local. >> > > I understand, but I want to install manually compiled packages > in /usr/local, so I set --prefix to that dir. I may have to give another > param to tell configure to search dependent libs? Anyway, it is not a big > problem. > > |
×
Want the latest updates on software, tech news, and AI?
Get latest updates about software, tech news, and AI from SourceForge directly in your inbox once a month.
Thanks for helping keep SourceForge clean.
X