visualpython-users

[Visualpython-users] last, desperate call for help

[Fionn B. <fi...@sp...>]

Hi all,


I have asked earlier, I do it again. We have a project here that needs 3d
output. the UI itself ist done in python-gtk.
We fail to manage starting vpython out of our gtk application. First the
two mainloops collided, so people told us to use python-gtk2 and threads.

So we switched the Ui to python-gtk2 and use threads. Similar to the
tk-demo thats coming with python-visual, we start python-visual in its own
thread.
The good news: no more mainloop clashing. The bad news: It segfaults
immediately, leaving loads of messages on the screen that look like this:

(<unknown>:13237): GLib-CRITICAL **: file ghash.c: line 225 (g_hash_table_lookup): assertion `hash_table != NULL' failed
(<unknown>:13237): Gtk-WARNING **: gtk_type_create(): unknown parent type `21'.
(<unknown>:13237): GLib-CRITICAL **: file ghash.c: line 225 (g_hash_table_lookup): assertion `hash_table != NULL' failed
(<unknown>:13237): GLib-CRITICAL **: file ghash.c: line 291 (g_hash_table_insert): assertion `hash_table != NULL' failed
(<unknown>:13237): GLib-CRITICAL **: file ghash.c: line 225 (g_hash_table_lookup): assertion `hash_table != NULL' failed
(<unknown>:13237): GLib-CRITICAL **: file ghash.c: line 291 (g_hash_table_insert): assertion `hash_table != NULL' failed
(<unknown>:13237): GLib-CRITICAL **: file ghash.c: line 225 (g_hash_table_lookup): assertion `hash_table != NULL' failed
(<unknown>:13237): GLib-CRITICAL **: file ghash.c: line 291 (g_hash_table_insert): assertion `hash_table != NULL' failed
(<unknown>:13237): GLib-CRITICAL **: file ghash.c: line 225 (g_hash_table_lookup): assertion `hash_table != NULL' failed
(<unknown>:13237): Gtk-WARNING **: gtk_type_create(): unknown parent type `21'.
(<unknown>:13237): Gtk-CRITICAL **: file gtktypeutils.c: line 337 (gtk_type_class): assertion `node != NULL' failed.
(<unknown>:13237): Gtk-WARNING **: gtk_arg_type_new(): argument class in "GtkContainer::border_width" is not in the `(null)' ancestry

...

Please, please, if anyone EVER managed to start vpython from a gtk gui or
has any tip on what might have gone wrong, send it!

Desperate,
		Fionn

Re: [Visualpython-users] last, desperate call for help

[Jonathan B. <jbr...@ea...>]

On Thu, 2004-09-09 at 13:04, Fionn Behrens wrote:
> Hi all,
> 
> 
> I have asked earlier, I do it again. We have a project here that needs 3d
> output. the UI itself ist done in python-gtk.
> We fail to manage starting vpython out of our gtk application. First the
> two mainloops collided, so people told us to use python-gtk2 and threads.
> 
> So we switched the Ui to python-gtk2 and use threads. Similar to the
> tk-demo thats coming with python-visual, we start python-visual in its own
> thread.
> The good news: no more mainloop clashing. The bad news: It segfaults
> immediately, leaving loads of messages on the screen that look like this:

Unfortunately, running Visual in a separate thread isn't enough - to be
totally safe right now, VPython has to be run in a separate process. 
The Gtk process would have to communicate with the Visual process using
a pipe, socket, or some other IPC mechanism.  However, there is hope.

Some significant work is being done to separate the OpenGL management
from the windowing system management, such that you will be able to
render a VPython scene into a toolkit-specific OpenGL window.  The
initial release of that functionality will include a PyGtkGLext-based
widget (gtkglext.sourceforge.net), since I am familiar with that
package.  It will also include the capability to do the same for other
toolkits in pure Python.

-Jonathan

Re: [Visualpython-users] last, desperate call for help

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

I'm not competent to comment directly, but could you perhaps handle this 
with more distance between your main application and the VPython viewer? 
For example, I have a Python program that helps with the logistics of 
grading student programs written in VPython (for an introductory physics 
course). The key statement is "exec(text)", where text was read from a 
file (exec is in the os module). At least as an experiment, you could 
write out information for VPython, do an exec of the VPython program, 
read up that information, and make the 3D display. Crude, inefficient, 
but it's hard to see how it could fail to work.

Bruce Sherwood

Fionn Behrens wrote:

>Hi all,
>
>
>I have asked earlier, I do it again. We have a project here that needs 3d
>output. the UI itself ist done in python-gtk.
>We fail to manage starting vpython out of our gtk application. First the
>two mainloops collided, so people told us to use python-gtk2 and threads.
>
>So we switched the Ui to python-gtk2 and use threads. Similar to the
>tk-demo thats coming with python-visual, we start python-visual in its own
>thread.
>The good news: no more mainloop clashing. The bad news: It segfaults
>immediately, leaving loads of messages on the screen that look like this:
>
>(<unknown>:13237): GLib-CRITICAL **: file ghash.c: line 225 (g_hash_table_lookup): assertion `hash_table != NULL' failed
>(<unknown>:13237): Gtk-WARNING **: gtk_type_create(): unknown parent type `21'.
>(<unknown>:13237): GLib-CRITICAL **: file ghash.c: line 225 (g_hash_table_lookup): assertion `hash_table != NULL' failed
>(<unknown>:13237): GLib-CRITICAL **: file ghash.c: line 291 (g_hash_table_insert): assertion `hash_table != NULL' failed
>(<unknown>:13237): GLib-CRITICAL **: file ghash.c: line 225 (g_hash_table_lookup): assertion `hash_table != NULL' failed
>(<unknown>:13237): GLib-CRITICAL **: file ghash.c: line 291 (g_hash_table_insert): assertion `hash_table != NULL' failed
>(<unknown>:13237): GLib-CRITICAL **: file ghash.c: line 225 (g_hash_table_lookup): assertion `hash_table != NULL' failed
>(<unknown>:13237): GLib-CRITICAL **: file ghash.c: line 291 (g_hash_table_insert): assertion `hash_table != NULL' failed
>(<unknown>:13237): GLib-CRITICAL **: file ghash.c: line 225 (g_hash_table_lookup): assertion `hash_table != NULL' failed
>(<unknown>:13237): Gtk-WARNING **: gtk_type_create(): unknown parent type `21'.
>(<unknown>:13237): Gtk-CRITICAL **: file gtktypeutils.c: line 337 (gtk_type_class): assertion `node != NULL' failed.
>(<unknown>:13237): Gtk-WARNING **: gtk_arg_type_new(): argument class in "GtkContainer::border_width" is not in the `(null)' ancestry
>
>...
>
>Please, please, if anyone EVER managed to start vpython from a gtk gui or
>has any tip on what might have gone wrong, send it!
>
>Desperate,
>		Fionn
>
>
>
>
>
>-------------------------------------------------------
>This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170
>Project Admins to receive an Apple iPod Mini FREE for your judgement on
>who ports your project to Linux PPC the best. Sponsored by IBM. 
>Deadline: Sept. 13. Go here: http://sf.net/ppc_contest.php
>_______________________________________________
>Visualpython-users mailing list
>Vis...@li...
>https://lists.sourceforge.net/lists/listinfo/visualpython-users
>  
>