visualpython-users

Re: [Visualpython-users] Linux installation

[Guy K. K. <guy...@au...>]

On Wed, 22 Dec 2010 13:18:22 Bruce Sherwood wrote:
> The appearance of "/usr/local" here is due to having specified
> PYTHON=/usr/bin/python2.6 and --prefix=/usr/local in the configure
> step for building Visual. On Ubuntu there exist
> /usr/lib/python2.6/dist-packages and also
> /usr/local/lib/python2.6/dist-packages,

Commonly user contributed installs should only go to e. g. /usr/local or /opt. 
Only distribution packages (including those from third parties) should install 
straight into the default file system hierarchy.

> and it is the latter folder where 3rd-party modules like Visual are
> apparently supposed to be installed.

Indeed. If I download the sources, compile and install, it should usually go 
into the /usr/local hierarchy. If there's something like a "make deb" or "make 
rpm" then that install can also go into the default system level hierarchy, as 
they're managed packages that can be properly maintained through the system 
provided packaging tools.

HTH,

Guy


-- 
Guy K. Kloss
School of Computing + Mathematical Sciences
Auckland University of Technology
Private Bag 92006, Auckland 1142
phone: +64 9 921 9999 ext. 5032
eMail: Guy...@au...

Re: [Visualpython-users] Linux installation

[Bruce S. <bas...@nc...>]

The appearance of "/usr/local" here is due to having specified
PYTHON=/usr/bin/python2.6 and --prefix=/usr/local in the configure
step for building Visual. On Ubuntu there exist
/usr/lib/python2.6/dist-packages and also
/usr/local/lib/python2.6/dist-packages, and it is the latter folder
where 3rd-party modules like Visual are apparently supposed to be
installed.

Bruce Sherwood

On Mon, Dec 20, 2010 at 10:19 PM, C Anthony Risinger <an...@ex...> wrote:
> On Mon, Dec 20, 2010 at 11:05 PM, Bruce Sherwood <bas...@nc...> wrote:
>> It does seem to be using the desired version of Python, as the full
>> path created is
>>
>> /usr/local/lib/python2.6/site-packages
>>
>> which contains the intended python2.6. I also tried replacing $PYTHON
>> with python$PYTHON_VERSION and still got site-packages. It's really
>> hard to understand how sysconfig.get_python_lib can act differently
>> within the autoconfigure environment.
>
> /usr/local?  shouldn't it be trying to link with the package manager's
> version?  at least, that's what your probably getting when you run
> `python` unless your doing something different.
>
> C Anthony
>

Re: [Visualpython-users] Linux installation

[C A. R. <an...@ex...>]

On Mon, Dec 20, 2010 at 11:05 PM, Bruce Sherwood <bas...@nc...> wrote:
> It does seem to be using the desired version of Python, as the full
> path created is
>
> /usr/local/lib/python2.6/site-packages
>
> which contains the intended python2.6. I also tried replacing $PYTHON
> with python$PYTHON_VERSION and still got site-packages. It's really
> hard to understand how sysconfig.get_python_lib can act differently
> within the autoconfigure environment.

/usr/local?  shouldn't it be trying to link with the package manager's
version?  at least, that's what your probably getting when you run
`python` unless your doing something different.

C Anthony

Re: [Visualpython-users] Linux installation

[Jacob S. <sch...@ro...>]

Bruce,

Ubuntu Lucid Lynx 10.04
=======================
Python 2.6.5 installed by default
Python 3.1.2 installed from repositories


I find this code in /usr/lib/python3.1/distutils/sysconfig.py
under the function get_python_lib



    if os.name == "posix":
        libpython = os.path.join(prefix,
                                 "lib", "python" + get_python_version())
        if standard_lib:
            return libpython
        elif is_default_prefix and 'PYTHONUSERBASE' not in os.environ
and 'real_prefix' not in sys.__dict__:
            return os.path.join(libpython, "dist-packages")
        else:
            return os.path.join(libpython, "site-packages")



It appears to suggest that the choice is made due to the elif line,
which appears to automatically choose 'site-packages' if you send in a
particular prefix.

Also, I would like to inform you that in my installation (Python 3.1.2),
site-packages is not on the search path, and the folder does not even
exist!

P.S. - Same result as above for my Python 2.6.5 installation

HTH,
Jacob Schmidt

On Mon, 2010-12-20 at 21:35 -0700, Bruce Sherwood wrote:
> I wonder whether someone on the list can help me with a puzzle having
> to do with installing on Linux.
> 
> In acinclude.m4 is the following fragment which attempts to identify
> where to install modules such as Visual:
> 
> [am_cv_python_pythondir=`$PYTHON -c "from distutils import sysconfig;
> print(sysconfig.get_python_lib(0,0,prefix='$PYTHON_PREFIX').replace('\\\\\','/'))"
> 2>/dev/null ||
> echo "$PYTHON_PREFIX/lib/python$PYTHON_VERSION/dist-packages"`]
> 
> When autogen.sh and configure are run before doing a make to build
> Visual, this fragment generates a reference to site-packages, which is
> unfortunate, because on Ubuntu with Python 2.x site-packages is not
> normally on the module search path.
> 
> If on the other hand I execute the following I get a reference to
> dist-packages, as expected and desired:
> 
> from distutils import sysconfig
> print(sysconfig.get_python_lib(0,0,prefix='$PYTHON_PREFIX')
> 
> I can't understand why I'm getting a reference to site-packages in the
> autoconfigure machinery. I'm reluctant to hand-code "dist-packages"
> because for all I know other Linux distributions use site-packages and
> have site-packages on the module search path. (As an aside, I'm
> annoyed that Ubuntu has left site-packages off the search path, though
> I notice that it seems to be back on the search path for Python 3.)
> 
> Bruce Sherwood
> 
> ------------------------------------------------------------------------------
> Lotusphere 2011
> Register now for Lotusphere 2011 and learn how
> to connect the dots, take your collaborative environment
> to the next level, and enter the era of Social Business.
> http://p.sf.net/sfu/lotusphere-d2d
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users

Re: [Visualpython-users] Linux installation

[Jacob S. <sch...@ro...>]

I should clarify that both versions of my pythons return site-packages
when run from the command prompt

Jacob


On Mon, 2010-12-20 at 22:05 -0700, Bruce Sherwood wrote:
> It does seem to be using the desired version of Python, as the full
> path created is
> 
> /usr/local/lib/python2.6/site-packages
> 
> which contains the intended python2.6. I also tried replacing $PYTHON
> with python$PYTHON_VERSION and still got site-packages. It's really
> hard to understand how sysconfig.get_python_lib can act differently
> within the autoconfigure environment.
> 
> Bruce Sherwood
> 
> On Mon, Dec 20, 2010 at 9:53 PM, C Anthony Risinger <an...@ex...> wrote:
> > On Mon, Dec 20, 2010 at 10:35 PM, Bruce Sherwood <bas...@nc...> wrote:
> >> I wonder whether someone on the list can help me with a puzzle having
> >> to do with installing on Linux.
> >>
> >> In acinclude.m4 is the following fragment which attempts to identify
> >> where to install modules such as Visual:
> >>
> >> [am_cv_python_pythondir=`$PYTHON -c "from distutils import sysconfig;
> >> print(sysconfig.get_python_lib(0,0,prefix='$PYTHON_PREFIX').replace('\\\\\','/'))"
> >> 2>/dev/null ||
> >> echo "$PYTHON_PREFIX/lib/python$PYTHON_VERSION/dist-packages"`]
> >>
> >> When autogen.sh and configure are run before doing a make to build
> >> Visual, this fragment generates a reference to site-packages, which is
> >> unfortunate, because on Ubuntu with Python 2.x site-packages is not
> >> normally on the module search path.
> >>
> >> If on the other hand I execute the following I get a reference to
> >> dist-packages, as expected and desired:
> >>
> >> from distutils import sysconfig
> >> print(sysconfig.get_python_lib(0,0,prefix='$PYTHON_PREFIX')
> >>
> >> I can't understand why I'm getting a reference to site-packages in the
> >> autoconfigure machinery. I'm reluctant to hand-code "dist-packages"
> >> because for all I know other Linux distributions use site-packages and
> >> have site-packages on the module search path. (As an aside, I'm
> >> annoyed that Ubuntu has left site-packages off the search path, though
> >> I notice that it seems to be back on the search path for Python 3.)
> >
> > this is only for ubuntu, right (appears that way)?
> >
> > i don't run ubuntu, so i can't really test, but are you sure autotools
> > is picking up the same version of python that you get when you just
> > run `python` in the terminal?
> >
> > i use archlinux, and recently we switched to py3k by default... so
> > i've had to deal with problems like that, and this feels like one of
> > those problems.  otherwise, there is something different in the
> > environments between your terminal and autotools, but i'm not sure how
> > to dump the ENV from within the macro.
> >
> > C Anthony
> >
> 
> ------------------------------------------------------------------------------
> Lotusphere 2011
> Register now for Lotusphere 2011 and learn how
> to connect the dots, take your collaborative environment
> to the next level, and enter the era of Social Business.
> http://p.sf.net/sfu/lotusphere-d2d
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users

Re: [Visualpython-users] Linux installation

[Bruce S. <bas...@nc...>]

Thanks much for the detailed information, though I don't understand
why this would function differently within and without the
autoconfigure environment.

As for site-packages being on the Python 3.1 search path, I see that I
had forgotten that I had put a .pth file in
/usr/local/python3.1/dist-packages to specify site-packages. Sorry for
the confusion (and sorry that Ubuntu has caused a lot of people grief
by not including site-packages, as judged from chatter on the web). An
amusing aside:

I chose the menu option System > About Ubuntu and got the following
remarkable message from the future:

You are using Ubuntu 11.04
                - the Natty Narwhal - released in April 2011 and
supported until October 2012.

However, I believe I'm using 10.10.
	
Bruce Sherwood

On Mon, Dec 20, 2010 at 10:06 PM, Jacob Schmidt
<sch...@ro...> wrote:
> Bruce,
>
> Ubuntu Lucid Lynx 10.04
> =======================
> Python 2.6.5 installed by default
> Python 3.1.2 installed from repositories
>
>
> I find this code in /usr/lib/python3.1/distutils/sysconfig.py
> under the function get_python_lib
>
>
>
>    if os.name == "posix":
>        libpython = os.path.join(prefix,
>                                 "lib", "python" + get_python_version())
>        if standard_lib:
>            return libpython
>        elif is_default_prefix and 'PYTHONUSERBASE' not in os.environ
> and 'real_prefix' not in sys.__dict__:
>            return os.path.join(libpython, "dist-packages")
>        else:
>            return os.path.join(libpython, "site-packages")
>
>
>
> It appears to suggest that the choice is made due to the elif line,
> which appears to automatically choose 'site-packages' if you send in a
> particular prefix.
>
> Also, I would like to inform you that in my installation (Python 3.1.2),
> site-packages is not on the search path, and the folder does not even
> exist!
>
> P.S. - Same result as above for my Python 2.6.5 installation
>
> HTH,
> Jacob Schmidt
>
> On Mon, 2010-12-20 at 21:35 -0700, Bruce Sherwood wrote:
>> I wonder whether someone on the list can help me with a puzzle having
>> to do with installing on Linux.
>>
>> In acinclude.m4 is the following fragment which attempts to identify
>> where to install modules such as Visual:
>>
>> [am_cv_python_pythondir=`$PYTHON -c "from distutils import sysconfig;
>> print(sysconfig.get_python_lib(0,0,prefix='$PYTHON_PREFIX').replace('\\\\\','/'))"
>> 2>/dev/null ||
>> echo "$PYTHON_PREFIX/lib/python$PYTHON_VERSION/dist-packages"`]
>>
>> When autogen.sh and configure are run before doing a make to build
>> Visual, this fragment generates a reference to site-packages, which is
>> unfortunate, because on Ubuntu with Python 2.x site-packages is not
>> normally on the module search path.
>>
>> If on the other hand I execute the following I get a reference to
>> dist-packages, as expected and desired:
>>
>> from distutils import sysconfig
>> print(sysconfig.get_python_lib(0,0,prefix='$PYTHON_PREFIX')
>>
>> I can't understand why I'm getting a reference to site-packages in the
>> autoconfigure machinery. I'm reluctant to hand-code "dist-packages"
>> because for all I know other Linux distributions use site-packages and
>> have site-packages on the module search path. (As an aside, I'm
>> annoyed that Ubuntu has left site-packages off the search path, though
>> I notice that it seems to be back on the search path for Python 3.)
>>
>> Bruce Sherwood
>>
>> ------------------------------------------------------------------------------
>> Lotusphere 2011
>> Register now for Lotusphere 2011 and learn how
>> to connect the dots, take your collaborative environment
>> to the next level, and enter the era of Social Business.
>> http://p.sf.net/sfu/lotusphere-d2d
>> _______________________________________________
>> Visualpython-users mailing list
>> Vis...@li...
>> https://lists.sourceforge.net/lists/listinfo/visualpython-users
>
>
>
>
>

Re: [Visualpython-users] Linux installation

[Bruce S. <bas...@nc...>]

It does seem to be using the desired version of Python, as the full
path created is

/usr/local/lib/python2.6/site-packages

which contains the intended python2.6. I also tried replacing $PYTHON
with python$PYTHON_VERSION and still got site-packages. It's really
hard to understand how sysconfig.get_python_lib can act differently
within the autoconfigure environment.

Bruce Sherwood

On Mon, Dec 20, 2010 at 9:53 PM, C Anthony Risinger <an...@ex...> wrote:
> On Mon, Dec 20, 2010 at 10:35 PM, Bruce Sherwood <bas...@nc...> wrote:
>> I wonder whether someone on the list can help me with a puzzle having
>> to do with installing on Linux.
>>
>> In acinclude.m4 is the following fragment which attempts to identify
>> where to install modules such as Visual:
>>
>> [am_cv_python_pythondir=`$PYTHON -c "from distutils import sysconfig;
>> print(sysconfig.get_python_lib(0,0,prefix='$PYTHON_PREFIX').replace('\\\\\','/'))"
>> 2>/dev/null ||
>> echo "$PYTHON_PREFIX/lib/python$PYTHON_VERSION/dist-packages"`]
>>
>> When autogen.sh and configure are run before doing a make to build
>> Visual, this fragment generates a reference to site-packages, which is
>> unfortunate, because on Ubuntu with Python 2.x site-packages is not
>> normally on the module search path.
>>
>> If on the other hand I execute the following I get a reference to
>> dist-packages, as expected and desired:
>>
>> from distutils import sysconfig
>> print(sysconfig.get_python_lib(0,0,prefix='$PYTHON_PREFIX')
>>
>> I can't understand why I'm getting a reference to site-packages in the
>> autoconfigure machinery. I'm reluctant to hand-code "dist-packages"
>> because for all I know other Linux distributions use site-packages and
>> have site-packages on the module search path. (As an aside, I'm
>> annoyed that Ubuntu has left site-packages off the search path, though
>> I notice that it seems to be back on the search path for Python 3.)
>
> this is only for ubuntu, right (appears that way)?
>
> i don't run ubuntu, so i can't really test, but are you sure autotools
> is picking up the same version of python that you get when you just
> run `python` in the terminal?
>
> i use archlinux, and recently we switched to py3k by default... so
> i've had to deal with problems like that, and this feels like one of
> those problems.  otherwise, there is something different in the
> environments between your terminal and autotools, but i'm not sure how
> to dump the ENV from within the macro.
>
> C Anthony
>

Re: [Visualpython-users] Linux installation

[C A. R. <an...@ex...>]

On Mon, Dec 20, 2010 at 10:35 PM, Bruce Sherwood <bas...@nc...> wrote:
> I wonder whether someone on the list can help me with a puzzle having
> to do with installing on Linux.
>
> In acinclude.m4 is the following fragment which attempts to identify
> where to install modules such as Visual:
>
> [am_cv_python_pythondir=`$PYTHON -c "from distutils import sysconfig;
> print(sysconfig.get_python_lib(0,0,prefix='$PYTHON_PREFIX').replace('\\\\\','/'))"
> 2>/dev/null ||
> echo "$PYTHON_PREFIX/lib/python$PYTHON_VERSION/dist-packages"`]
>
> When autogen.sh and configure are run before doing a make to build
> Visual, this fragment generates a reference to site-packages, which is
> unfortunate, because on Ubuntu with Python 2.x site-packages is not
> normally on the module search path.
>
> If on the other hand I execute the following I get a reference to
> dist-packages, as expected and desired:
>
> from distutils import sysconfig
> print(sysconfig.get_python_lib(0,0,prefix='$PYTHON_PREFIX')
>
> I can't understand why I'm getting a reference to site-packages in the
> autoconfigure machinery. I'm reluctant to hand-code "dist-packages"
> because for all I know other Linux distributions use site-packages and
> have site-packages on the module search path. (As an aside, I'm
> annoyed that Ubuntu has left site-packages off the search path, though
> I notice that it seems to be back on the search path for Python 3.)

this is only for ubuntu, right (appears that way)?

i don't run ubuntu, so i can't really test, but are you sure autotools
is picking up the same version of python that you get when you just
run `python` in the terminal?

i use archlinux, and recently we switched to py3k by default... so
i've had to deal with problems like that, and this feels like one of
those problems.  otherwise, there is something different in the
environments between your terminal and autotools, but i'm not sure how
to dump the ENV from within the macro.

C Anthony

[Visualpython-users] Linux installation

[Bruce S. <bas...@nc...>]

I wonder whether someone on the list can help me with a puzzle having
to do with installing on Linux.

In acinclude.m4 is the following fragment which attempts to identify
where to install modules such as Visual:

[am_cv_python_pythondir=`$PYTHON -c "from distutils import sysconfig;
print(sysconfig.get_python_lib(0,0,prefix='$PYTHON_PREFIX').replace('\\\\\','/'))"
2>/dev/null ||
echo "$PYTHON_PREFIX/lib/python$PYTHON_VERSION/dist-packages"`]

When autogen.sh and configure are run before doing a make to build
Visual, this fragment generates a reference to site-packages, which is
unfortunate, because on Ubuntu with Python 2.x site-packages is not
normally on the module search path.

If on the other hand I execute the following I get a reference to
dist-packages, as expected and desired:

from distutils import sysconfig
print(sysconfig.get_python_lib(0,0,prefix='$PYTHON_PREFIX')

I can't understand why I'm getting a reference to site-packages in the
autoconfigure machinery. I'm reluctant to hand-code "dist-packages"
because for all I know other Linux distributions use site-packages and
have site-packages on the module search path. (As an aside, I'm
annoyed that Ubuntu has left site-packages off the search path, though
I notice that it seems to be back on the search path for Python 3.)

Bruce Sherwood

Re: [Visualpython-users] segfault with rotate() using visual 5.4 on Debian i386

[Bruce S. <bas...@nc...>]

When I run your program on Windows I get this:

Traceback (most recent call last):
  File "Untitled", line 4
    b.rotate(pi/4)
ValueError: primitive.rotate(): angle of rotation must be specified.

The rotate statement should look like this  b.rotate(angle=pi/4)

The C++ code at wrap_primitive.cpp, line 48, is this:

throw std::invalid_argument(
            "primitive.rotate(): angle of rotation must be specified.");

I don't know why Debian would segfault rather than giving an error.
However, I note that translate_exception shows up in your error
report, and in INSTALL.txt there is the following comment (though if
this were the problem I would expect the compilation to fail):

If you get this error in compiling cvisualmodule.cpp:
		
    In file included from /usr/include/boost/python/exception_translator.hpp:12,
    from ../../vpython-core2/src/python/cvisualmodule.cpp:12:
    /usr/include/boost/python/detail/translate_exception.hpp:34:
	      error: expected nested-name-specifier before "add_reference"

	You need to make this change:
	
	In /usr/include (probably), change /boost/python/detail/translate_exception.hpp
	=========================================================
	--- boost/python/detail/translate_exception.hpp (revision 50228)
	+++ boost/python/detail/translate_exception.hpp (working copy)
	@@ -9,6 +9,7 @@

	# include <boost/call_traits.hpp>
 	# include <boost/type_traits/add_const.hpp>
	+# include <boost/type_traits/add_reference.hpp>

	# include <boost/function/function0.hpp>
	--------------------------------------
	(That is, add "# include <boost/type_traits/add_reference.hpp>" after the
	statement "# include <boost/type_traits/add_const.hpp>".)

Bruce Sherwood

[Visualpython-users] segfault with rotate() using visual 5.4 on Debian i386

[Sanjoy M. <sa...@ol...>]

I compiled visual 5.4 from source for my laptop, a Thinkpad T60 that
runs a Debian testing/unstable mix.  The following short program gives a
segfault:

  from visual import *
  b=box()
  scene.autoscale=0
  b.rotate(pi/4)

It's using the boost 1.42.0 libraries and python 2.6.6.

Here's the backtrace when it is run under gdb.  Any other debugging
steps to do?

$ LIBGL_ALWAYS_SOFTWARE=1 PYTHONPATH="/usr/local/lib/python2.6/site-packages" gdb --args python "/home/sanjoy/admin/bug/vpython/test.py"
(gdb) run
Starting program: /usr/bin/python /home/sanjoy/admin/bug/vpython/test.py
[Thread debugging using libthread_db enabled]
[New Thread 0xb533cb70 (LWP 32322)]

Program received signal SIGSEGV, Segmentation fault.
0xb676cbb7 in __cxa_allocate_exception () from /usr/lib/libstdc++.so.6
(gdb) bt
#0  0xb676cbb7 in __cxa_allocate_exception () from /usr/lib/libstdc++.so.6
#1  0xb7a269f7 in cvisual::py_rotate<cvisual::primitive> (args=..., kwargs=DWARF-2 expression error: DW_OP_reg operations must be used either alone or in conjuction with DW_OP_piece or DW_OP_bit_piece.
)
    at ../../visual-5.4_release/src/python/wrap_primitive.cpp:48
#2  0xb7a2665d in operator() (this=0x83067a8, args=0x84a7dac, kw=0x0)
    at /usr/include/boost/python/raw_function.hpp:36
#3  boost::python::objects::full_py_function_impl<boost::python::detail::raw_dispatcher<boost::python::api::object (*)(boost::python::tuple, boost::python::dict)>, boost::mpl::vector1<PyObject*> >::operator() (this=0x83067a8, 
    args=0x84a7dac, kw=0x0)
    at /usr/include/boost/python/object/py_function.hpp:94
#4  0xb67f701f in boost::python::objects::function::call(_object*, _object*) const () from /usr/lib/libboost_python-py26.so.1.42.0
#5  0xb67f7257 in ?? () from /usr/lib/libboost_python-py26.so.1.42.0
#6  0xb67fe532 in boost::python::detail::exception_handler::operator()(boost::function0<void> const&) const () from /usr/lib/libboost_python-py26.so.1.42.0
#7  0xb79dbd82 in operator() (function_obj_ptr=..., a0=..., a1=...)
    at /usr/include/boost/python/detail/translate_exception.hpp:48
#8  operator()<bool, boost::python::detail::translate_exception<std::runtime_error, void (*)(std::runtime_error)>, boost::_bi::list2<const boost::python::detail::exception_handler&, const boost::function0<void>&> > (
    function_obj_ptr=..., a0=..., a1=...)
    at /usr/include/boost/bind/bind.hpp:382
#9  operator()<boost::python::detail::exception_handler, boost::function0<void> > (function_obj_ptr=..., a0=..., a1=...)
    at /usr/include/boost/bind/bind_template.hpp:102
#10 boost::detail::function::function_obj_invoker2<boost::_bi::bind_t<bool, boost::python::detail::translate_exception<std::runtime_error, void (*)(std::runtime_error)>, boost::_bi::list3<boost::arg<1>, boost::arg<2>, boost::_bi::value<void (*)(std::runtime_error)> > >, bool, boost::python::detail::exception_handler const&, boost::function0<void> const&>::invoke (function_obj_ptr=..., 
    a0=..., a1=...) at /usr/include/boost/function/function_template.hpp:132
#11 0xb67fe511 in boost::python::detail::exception_handler::operator()(boost::function0<void> const&) const () from /usr/lib/libboost_python-py26.so.1.42.0
#12 0xb79dbb22 in operator() (function_obj_ptr=..., a0=..., a1=...)
    at /usr/include/boost/python/detail/translate_exception.hpp:48
#13 operator()<bool, boost::python::detail::translate_exception<std::invalid_argument, void (*)(std::invalid_argument)>, boost::_bi::list2<const boost::python::detail::exception_handler&, const boost::function0<void>&> > (
    function_obj_ptr=..., a0=..., a1=...)
    at /usr/include/boost/bind/bind.hpp:382
#14 operator()<boost::python::detail::exception_handler, boost::function0<void> > (function_obj_ptr=..., a0=..., a1=...)
    at /usr/include/boost/bind/bind_template.hpp:102
#15 boost::detail::function::function_obj_invoker2<boost::_bi::bind_t<bool, boost::python::detail::translate_exception<std::invalid_argument, void (*)(std::invalid_argument)>, boost::_bi::list3<boost::arg<1>, boost::arg<2>, boost::_bi::value<void (*)(std::invalid_argument)> > >, bool, boost::python::detail::exception_handler const&, boost::function0<void> const&>::invoke (
    function_obj_ptr=..., a0=..., a1=...)
    at /usr/include/boost/function/function_template.hpp:132
#16 0xb67fe511 in boost::python::detail::exception_handler::operator()(boost::function0<void> const&) const () from /usr/lib/libboost_python-py26.so.1.42.0
#17 0xb79dba72 in operator() (function_obj_ptr=..., a0=..., a1=...)
    at /usr/include/boost/python/detail/translate_exception.hpp:48
#18 operator()<bool, boost::python::detail::translate_exception<std::out_of_range, void (*)(std::out_of_range)>, boost::_bi::list2<const boost::python::detail::exception_handler&, const boost::function0<void>&> > (function_obj_ptr=..., 
    a0=..., a1=...) at /usr/include/boost/bind/bind.hpp:382
#19 operator()<boost::python::detail::exception_handler, boost::function0<void> > (function_obj_ptr=..., a0=..., a1=...)
    at /usr/include/boost/bind/bind_template.hpp:102
#20 boost::detail::function::function_obj_invoker2<boost::_bi::bind_t<bool, boost::python::detail::translate_exception<std::out_of_range, void (*)(std::out_of_range)>, boost::_bi::list3<boost::arg<1>, boost::arg<2>, boost::_bi::value<void (*)(std::out_of_range)> > >, bool, boost::python::detail::exception_handler const&, boost::function0<void> const&>::invoke (function_obj_ptr=..., a0=..., 
    a1=...) at /usr/include/boost/function/function_template.hpp:132
#21 0xb67fe303 in boost::python::handle_exception_impl(boost::function0<void>)
    () from /usr/lib/libboost_python-py26.so.1.42.0
#22 0xb67f3da6 in ?? () from /usr/lib/libboost_python-py26.so.1.42.0
#23 0x0806232a in PyObject_Call ()
#24 0x080e016b in PyEval_EvalFrameEx ()
#25 0x080e2507 in PyEval_EvalCodeEx ()
#26 0x080e2607 in PyEval_EvalCode ()
#27 0x080ffcbd in PyRun_FileExFlags ()
#28 0x080fff22 in PyRun_SimpleFileExFlags ()
#29 0x0805dd81 in Py_Main ()
#30 0x0805cf6b in main ()

Re: [Visualpython-users] More objects

[Bruce S. <bas...@nc...>]

Kadir Haldenbilen points out a failure of the imagination on my part.
An element of our plan is to be able to specify an array for the
radius attribute (or consider it a scale factor) so that the curve's
cross section can vary along the curve. This will make it possible to
form frustums, and partial spheres.

Bruce Sherwood

On Sun, Dec 19, 2010 at 10:40 AM, Bruce Sherwood <bas...@nc...> wrote:
> Those objects are built using faces, which currently support materials
> but not opacity. You might consider adding a materials option to those
> objects yourself.
>
> Kadir Haldenbilen, whose work with the Polygon, FontTools, and
> ttfquery modules led to the 3D text object, has done further work on
> 2D "constructive geometry" using the Polygon tool, which makes it
> possible to add and subtract areas to create arbitrarily complex 2D
> figures such as a square with round holes.
>
> He and I hope to incorporate this machinery into VPython by allowing
> you to specify one of these 2D areas as a cross section for the curve
> object. This would give you hollow tubes and much much more, but it
> wouldn't give you frustums and partial spheres (which would require 3D
> constructive geometry).
>
> Bruce Sherwood
>
> On Sun, Dec 19, 2010 at 9:33 AM, Poul Riis <Pou...@sk...> wrote:
>> I'm still hoping that someone could find the time to update Thom Ives'
>> moreobjects.py so that the tube, the frustum, and the spherepart could be
>> used with the materials and the opacity options - or even better include
>> these objects as standard objects in VPython.
>> Furthermore, I would suggest a ring section (bent tube) as a new object (for
>> connecting tubes, for instance).
>> Poul Riis
>> _______________________________________________
>> Visualpython-users mailing list
>> Vis...@li...
>> https://lists.sourceforge.net/lists/listinfo/visualpython-users
>>
>>
>

Re: [Visualpython-users] More objects

[Bruce S. <bas...@nc...>]

Those objects are built using faces, which currently support materials
but not opacity. You might consider adding a materials option to those
objects yourself.

Kadir Haldenbilen, whose work with the Polygon, FontTools, and
ttfquery modules led to the 3D text object, has done further work on
2D "constructive geometry" using the Polygon tool, which makes it
possible to add and subtract areas to create arbitrarily complex 2D
figures such as a square with round holes.

He and I hope to incorporate this machinery into VPython by allowing
you to specify one of these 2D areas as a cross section for the curve
object. This would give you hollow tubes and much much more, but it
wouldn't give you frustums and partial spheres (which would require 3D
constructive geometry).

Bruce Sherwood

On Sun, Dec 19, 2010 at 9:33 AM, Poul Riis <Pou...@sk...> wrote:
> I'm still hoping that someone could find the time to update Thom Ives'
> moreobjects.py so that the tube, the frustum, and the spherepart could be
> used with the materials and the opacity options - or even better include
> these objects as standard objects in VPython.
> Furthermore, I would suggest a ring section (bent tube) as a new object (for
> connecting tubes, for instance).
> Poul Riis
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users
>
>

[Visualpython-users] More objects

[Poul R. <Pou...@sk...>]

I'm still hoping that someone could find the time to update Thom Ives'
moreobjects.py so that the tube, the frustum, and the spherepart could be
used with the materials and the opacity options - or even better include
these objects as standard objects in VPython.
Furthermore, I would suggest a ring section (bent tube) as a new object
(for connecting tubes, for instance).

Poul Riis

[Visualpython-users] Smooth or jagged lines and edges

[Bruce S. <bas...@nc...>]

In response to a question about jagged lines and edges in VPython
programs, I've added this to the Windows and Linux download pages:

If you see overly jagged lines and edges, you may be able to smooth
them by increasing "antialiasing" in the advanced section of your
graphics card settings, and/or choosing "quality" (appearance of
graphics) over "performance" (speed of graphics).

Bruce Sherwood

[Visualpython-users] problem compiling arraryprim

[Herbert W. <we...@su...>]

Hi,

I can't get arrayprim compiled when attempting to install
visual5.4 on opensuse 11.3 64bit. Configure runs well but
when it comes to compile arrayprim, the compilation is
terminated on 'Error 1'. Some time ago there was a similar
discussion on the list regarding MAC. Unfortunately, I was
not able to extract a suitable solution for my problem.

Thanks in advance for your help.
Herbert.

PS currently I do not have access to that 64bit machine. If you
need further info (log files etc) I will send them later.

[Visualpython-users] VPython 5.41 for Windows

[Bruce S. <bas...@nc...>]

Now available at vpython.org: VPython 5.41 for Windows, which fixes
bugs in active and passive stereo.

Bruce Sherwood

Re: [Visualpython-users] objects only appear as black

[Vijay S. <vij...@gm...>]

I can live with that materials issue for now, though I understand your point
about there possibly being other problems.  I looked in to the updating the
driver, and as far as I can tell, I have the newest driver for my hardware.
That'll probably change later on though, so I'll keep an eye out for an
update.

Thanks for the heads up.

On Tue, Nov 30, 2010 at 4:13 PM, Bruce Sherwood <bas...@nc...> wrote:

> It doesn't affect the speed of rendering. What it does is turn off the
> attempt to render materials, so that
>
>   box(color=color.orange, material=materials.wood)
>
> displays simply as an orange box, with no wood grain. The connection
> with my earlier suggestion to update the graphics driver is that
> graphics cards and drivers that cannot handle these materials may
> exhibit other problems as well.
>
> Bruce Sherwood
>
> On Tue, Nov 30, 2010 at 2:04 PM, Vijay Somers <vij...@gm...>
> wrote:
> > Thanks, that worked.  Does making this change lower my machine's
> rendering
> > performance while using vpython?
> >
> > On Tue, Nov 30, 2010 at 11:58 AM, Aaron Miller <ajm...@al...>
> wrote:
> >>
> >> Vijay,
> >>
> >> I have also seen this on older hardware (on Windows, at least).  The
> >> problem goes away when using the legacy shader rendering.  Make this
> change
> >> by editing the site_settings.py file (found at
> >> C:\Python27\Lib\site-packages\visual for my windows install) and
> >> uncommenting the line that reads display.enable_shaders = False.
> >>
> >> I hope it works for you!
> >>
> >> Aaron
>
>
> ------------------------------------------------------------------------------
> Increase Visibility of Your 3D Game App & Earn a Chance To Win $500!
> Tap into the largest installed PC base & get more eyes on your game by
> optimizing for Intel(R) Graphics Technology. Get started today with the
> Intel(R) Software Partner Program. Five $500 cash prizes are up for grabs.
> http://p.sf.net/sfu/intelisp-dev2dev
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users
>

Re: [Visualpython-users] objects only appear as black

[Bruce S. <bas...@nc...>]

It doesn't affect the speed of rendering. What it does is turn off the
attempt to render materials, so that

   box(color=color.orange, material=materials.wood)

displays simply as an orange box, with no wood grain. The connection
with my earlier suggestion to update the graphics driver is that
graphics cards and drivers that cannot handle these materials may
exhibit other problems as well.

Bruce Sherwood

On Tue, Nov 30, 2010 at 2:04 PM, Vijay Somers <vij...@gm...> wrote:
> Thanks, that worked.  Does making this change lower my machine's rendering
> performance while using vpython?
>
> On Tue, Nov 30, 2010 at 11:58 AM, Aaron Miller <ajm...@al...> wrote:
>>
>> Vijay,
>>
>> I have also seen this on older hardware (on Windows, at least).  The
>> problem goes away when using the legacy shader rendering.  Make this change
>> by editing the site_settings.py file (found at
>> C:\Python27\Lib\site-packages\visual for my windows install) and
>> uncommenting the line that reads display.enable_shaders = False.
>>
>> I hope it works for you!
>>
>> Aaron

Re: [Visualpython-users] objects only appear as black

[Vijay S. <vij...@gm...>]

Thanks, that worked.  Does making this change lower my machine's rendering
performance while using vpython?

On Tue, Nov 30, 2010 at 11:58 AM, Aaron Miller <ajm...@al...> wrote:

> Vijay,
>
> I have also seen this on older hardware (on Windows, at least).  The
> problem goes away when using the legacy shader rendering.  Make this change
> by editing the site_settings.py file (found at
> C:\Python27\Lib\site-packages\visual for my windows install) and
> uncommenting the line that reads display.enable_shaders = False.
>
> I hope it works for you!
>
> Aaron
>
> ------------------------------------------------------
> Aaron J. Miller, Ph.D.
> Department of Physics
> 328 Palenske Hall
> Albion College, Albion, MI 49224
>
>>
>> Message: 5
>> Date: Mon, 29 Nov 2010 22:20:19 -0700
>> From: Bruce Sherwood <bas...@nc...>
>> Subject: Re: [Visualpython-users] objects only appear as black
>>        silhouettes
>> To: vis...@li...
>> Message-ID:
>>        <AAN...@ma...<AANLkTin%2B6...@ma...>
>> >
>> Content-Type: text/plain; charset=ISO-8859-1
>>
>> The problem almost certainly is that you need to update your graphics
>> driver.
>>
>> Bruce Sherwood
>>
>> On Mon, Nov 29, 2010 at 7:58 PM, Vijay Somers <vij...@gm...>
>> wrote:
>> > Hi,
>> > I'm just trying to run the following simple bouncing ball example from
>> the
>> > vpython site:
>> >
>> > from visual import *
>> >
>> > floor = box (pos=(0,0,0), length=4, height=0.5, width=4,
>> color=color.blue)
>> > ball = sphere (pos=(0,4,0), radius=1, color=color.red)
>> > ball.velocity = vector(0,-1,0)
>> > dt = 0.01
>> >
>> > while 1:
>> > ????rate (100)
>> > ????ball.pos = ball.pos + ball.velocity*dt
>> > ????if ball.y < ball.radius:
>> > ????????ball.velocity.y = abs(ball.velocity.y)
>> > ????else:
>> > ????????ball.velocity.y = ball.velocity.y - 9.8*dt
>> >
>> > It runs without errors, but the floor and ball objects are black,
>> without
>> > any visible surface detail.? The silhouettes of the objects are visible
>> if I
>> > change the background colour to something other than black, but changes
>> to
>> > the colour of the objects don't affect anything.? I'm running this on a
>> > fresh install of Ubuntu 10.04, python 2.6.6, and it says I have version
>> > 1:5.12-1.1 of vpython.
>> >
>> > Any help would be appreciated,
>> > Thanks,
>> > Vijay
>> ___________________
>> > Visualpython-users mailing list
>> > Vis...@li...
>> > https://lists.sourceforge.net/lists/listinfo/visualpython-users
>> >
>> >
>>
>>
>
>
> ------------------------------------------------------------------------------
> Increase Visibility of Your 3D Game App & Earn a Chance To Win $500!
> Tap into the largest installed PC base & get more eyes on your game by
> optimizing for Intel(R) Graphics Technology. Get started today with the
> Intel(R) Software Partner Program. Five $500 cash prizes are up for grabs.
> http://p.sf.net/sfu/intelisp-dev2dev
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users
>
>

Re: [Visualpython-users] objects only appear as black

[Steve B. <sbe...@gm...>]

Thank you! Works like a charm. The Sun is now orange. Mars is properly red.
And Uranus and Neptune are properly blue.

In a manner of speaking, you've saved the Solar System! :-)

-Steve

Re: [Visualpython-users] objects only appear as black

[Aaron M. <ajm...@al...>]

Vijay,

I have also seen this on older hardware (on Windows, at least).  The problem
goes away when using the legacy shader rendering.  Make this change by
editing the site_settings.py file (found at
C:\Python27\Lib\site-packages\visual for my windows install) and
uncommenting the line that reads display.enable_shaders = False.

I hope it works for you!

Aaron

------------------------------------------------------
Aaron J. Miller, Ph.D.
Department of Physics
328 Palenske Hall
Albion College, Albion, MI 49224

>
> Message: 5
> Date: Mon, 29 Nov 2010 22:20:19 -0700
> From: Bruce Sherwood <bas...@nc...>
> Subject: Re: [Visualpython-users] objects only appear as black
>        silhouettes
> To: vis...@li...
> Message-ID:
>        <AAN...@ma...<AANLkTin%2B6...@ma...>
> >
> Content-Type: text/plain; charset=ISO-8859-1
>
> The problem almost certainly is that you need to update your graphics
> driver.
>
> Bruce Sherwood
>
> On Mon, Nov 29, 2010 at 7:58 PM, Vijay Somers <vij...@gm...>
> wrote:
> > Hi,
> > I'm just trying to run the following simple bouncing ball example from
> the
> > vpython site:
> >
> > from visual import *
> >
> > floor = box (pos=(0,0,0), length=4, height=0.5, width=4,
> color=color.blue)
> > ball = sphere (pos=(0,4,0), radius=1, color=color.red)
> > ball.velocity = vector(0,-1,0)
> > dt = 0.01
> >
> > while 1:
> > ????rate (100)
> > ????ball.pos = ball.pos + ball.velocity*dt
> > ????if ball.y < ball.radius:
> > ????????ball.velocity.y = abs(ball.velocity.y)
> > ????else:
> > ????????ball.velocity.y = ball.velocity.y - 9.8*dt
> >
> > It runs without errors, but the floor and ball objects are black, without
> > any visible surface detail.? The silhouettes of the objects are visible
> if I
> > change the background colour to something other than black, but changes
> to
> > the colour of the objects don't affect anything.? I'm running this on a
> > fresh install of Ubuntu 10.04, python 2.6.6, and it says I have version
> > 1:5.12-1.1 of vpython.
> >
> > Any help would be appreciated,
> > Thanks,
> > Vijay
> ___________________
> > Visualpython-users mailing list
> > Vis...@li...
> > https://lists.sourceforge.net/lists/listinfo/visualpython-users
> >
> >
>
>

Re: [Visualpython-users] objects only appear as black silhouettes

[Bruce S. <bas...@nc...>]

The problem almost certainly is that you need to update your graphics driver.

Bruce Sherwood

On Mon, Nov 29, 2010 at 7:58 PM, Vijay Somers <vij...@gm...> wrote:
> Hi,
> I'm just trying to run the following simple bouncing ball example from the
> vpython site:
>
> from visual import *
>
> floor = box (pos=(0,0,0), length=4, height=0.5, width=4, color=color.blue)
> ball = sphere (pos=(0,4,0), radius=1, color=color.red)
> ball.velocity = vector(0,-1,0)
> dt = 0.01
>
> while 1:
>     rate (100)
>     ball.pos = ball.pos + ball.velocity*dt
>     if ball.y < ball.radius:
>         ball.velocity.y = abs(ball.velocity.y)
>     else:
>         ball.velocity.y = ball.velocity.y - 9.8*dt
>
> It runs without errors, but the floor and ball objects are black, without
> any visible surface detail.  The silhouettes of the objects are visible if I
> change the background colour to something other than black, but changes to
> the colour of the objects don't affect anything.  I'm running this on a
> fresh install of Ubuntu 10.04, python 2.6.6, and it says I have version
> 1:5.12-1.1 of vpython.
>
> Any help would be appreciated,
> Thanks,
> Vijay
___________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users
>
>

[Visualpython-users] objects only appear as black silhouettes

[Vijay S. <vij...@gm...>]

Hi,
I'm just trying to run the following simple bouncing ball example from the
vpython site:

*from visual import **

*floor = box (pos=(0,0,0), length=4, height=0.5, width=4, color=color.blue)*
*ball = sphere (pos=(0,4,0), radius=1, color=color.red)*
*ball.velocity = vector(0,-1,0)*
*dt = 0.01*

*while 1:*
*    rate (100)*
*    ball.pos = ball.pos + ball.velocity*dt*
*    if ball.y < ball.radius:*
*        ball.velocity.y = abs(ball.velocity.y*)
*    else:*
*        ball.velocity.y = ball.velocity.y - 9.8*dt*


It runs without errors, but the floor and ball objects are black, without
any visible surface detail.  The silhouettes of the objects are visible if I
change the background colour to something other than black, but changes to
the colour of the objects don't affect anything.  I'm running this on a
fresh install of Ubuntu 10.04, python 2.6.6, and it says I have version
1:5.12-1.1 of vpython.

Any help would be appreciated,
Thanks,
Vijay

[Visualpython-users] VPython 5.4 for Python 2.6

[Bruce S. <bas...@nc...>]

Now available at vpython.org is a VPython 5.4 package for Python 2.6
for Windows and Mac, which includes numpy 1.5.1. Wihtout this, anyone
who wanted to use both VPython and scipy had no solution, because
there was no scipy module for Python 2.7 or Python 3.1, and the
existing VPython 5.32 for Python 2.6 included numpy 1.3, whereas scipy
now requires at least numpy 1.4.

The modularity of Python modules isn't as modular as one would like.
When VPython 5.32 was released for Python 2.6, the current numpy was
1.3. Then later numpy was upgraded to 1.4, and scipy was upgraded to
require numpy 1.4, leaving a dilemma. Sigh.

Bruce Sherwood