visualpython-users

Re: [Visualpython-users] vpython and wxpython integration

[CL <clc...@gm...>]

Hi Shef

I've uploaded my version of cvisual.pyd for your testing, if you are
interested. You can get it from http://code.google.com/p/vpythonex1

The cvisual.pyd-fix3 contains 3 fixes. Download and overwrite your
C:\Python25\Lib\site-packages\visual\cvisual.pyd

The three fixes are:
1. Fix the problem of my video card opengl extension library entry. It
is not compatible with vpython 5.0. This fix shall not affecting you,
I believe.

2. Fix the crash when opacity changed from 1.0 to below 1.0

3. The fix of mouse cursor positioning problem caused by re-parenting
in the docking situation.

CL

Re: [Visualpython-users] vpython and wxpython integration

[CL <clc...@gm...>]

Hi Stef

I've uploaded my version of cvisual.pyd for your testing, if you are
interested. You can get it from http://code.google.com/p/vpythonex1

The cvisual.pyd-fix3 contains 3 fixes. Download and overwrite your
C:\Python25\Lib\site-packages\visual\cvisual.pyd

The three fixes are:
1. Fix the problem of my video card opengl extension library entry. It
is not compatible with vpython 5.0. This fix shall not affecting you,
I believe.

2. Fix the crash when opacity changed from 1.0 to below 1.0

3. The fix of mouse cursor positioning problem caused by re-parenting
in the docking situation.

CL

[Visualpython-users] keyboard snag

[symion <sy...@pr...>]

I have been looking workarounds and found that by including opacity 
during a box definition, with any value Except 1.0 allows opacity to work!
Also, if you have nested frames then a pause before connecting the main 
frame from the sub frames works!

Another problem that has just popped up.
Any keyboard input that requires holding down a key for long periods 
seems to cause some programs to freeze, but not all!

Here are 3 examples, they all use cursor keys to move or spin objects.

result-1.py 
<http://home.primusonline.com.au/knoware/python/result-1.py>  This one 
crashes easily.

Spincube.py 
<http://home.primusonline.com.au/knoware/python/Spincube.py>  This one 
does not crash at all (at least on my machine)

Wiregrid-0.py 
<http://home.primusonline.com.au/knoware/python/Wiregrid-0.py>  This one 
crashes after a while. You have to move the cursor a fair way before it 
locks up.


It acts like a keyboard buffer overflow, but that's just a guess.

Also, vpython programs seem to take a long time to exit. I'm on a PC and 
with Task Manager set to monitor processes, it takes as much as 11 
seconds for the pythonw.exe to be removed from the list. Now maybe it's 
just my slow system, (I'm running Vista Home Edition) but if I try 
running another script before the old one has been cleared out then the 
threads get jumbled and I either get a 'blocked by firewall' message or 
a 'bad socket' message.

Symion

Re: [Visualpython-users] vpython and wxpython integration

[Stef M. <s.m...@ru...>]

CL wrote:
> Hi Shef
>
> Now I understand more about the problem you described.
>
> Actually vPython uses client windows coordinates in its mouse
> management, but it has the assumption that it is the top-level window
> when restoring the mouse cursor position after dragging. Somehow it
> will cause a crash in some cases. I've fixed that portion and I am not
> able to crash it any more. (May be I am not trying hard enough 8-)
>
>   
that's great news !!
And I would love to try it,
but the code below looks like C to me,
and that's beyond my knowledge :-(

cheers,
Stef
> In windisplay.cpp
>
> LRESULT
> display::on_mouse( WPARAM wParam, LPARAM lParam)
> ...
> if (mouse.is_mouse_locked() && ( mouse_x < 20 || mouse_x > view_width
> - 20 || mouse_y < 20 || mouse_y > view_height - 20 ) ) {
> 		mouse.report_setcursor( view_width/2, view_height/2 );
> 		SetCursorPos( view_x + view_width/2, view_y + view_height/2 );
> 	}
> 	if (was_locked && !mouse.is_mouse_locked()) {
> 		mouse.report_setcursor( old_x, old_y );
> 		SetCursorPos( view_x + old_x, view_y + old_y );
> 	}
> ...
>
>
> Both SetCursorPos call will move the mouse to an incorrect coordinate.
> It is because WM_MOVE message is not sending to vPython any more after
> it is reparented. (also, may be window_x and window_y shall be used
> instead of view_x, view_y).
>
> Changing it to:
>
> 	if (mouse.is_mouse_locked() && ( mouse_x < 20 || mouse_x > view_width
> - 20 || mouse_y < 20 || mouse_y > view_height - 20 ) ) {
> 		mouse.report_setcursor( view_width/2, view_height/2 );
> 		//SetCursorPos( view_x + view_width/2, view_y + view_height/2 );
> 		POINT pt;
> 		pt.x = pt.y = 0;
> 		ClientToScreen(widget_handle, &pt);
> 		SetCursorPos( pt.x + view_width/2, pt.y + view_height/2 );
> 	}
> 	if (was_locked && !mouse.is_mouse_locked()) {
> 		mouse.report_setcursor( old_x, old_y );
> 		//SetCursorPos( view_x + old_x, view_y + old_y );
> 		POINT pt;
> 		pt.x = pt.y = 0;
> 		ClientToScreen(widget_handle, &pt);
> 		SetCursorPos( pt.x + old_x,  pt.y + old_y );
> 	}
>
> Now the mouse cursor is correctly positioned after dragging, no matter
> if it has been reparented or not. Seems no more crashes now.
>
> CL
>
>   
>


Het UMC St Radboud staat geregistreerd bij de Kamer van Koophandel in het handelsregister onder nummer 41055629.
The Radboud University Nijmegen Medical Centre is listed in the Commercial Register of the Chamber of Commerce under file number 41055629.

Re: [Visualpython-users] vpython and wxpython integration

[CL <clc...@gm...>]

Hi Shef

Now I understand more about the problem you described.

Actually vPython uses client windows coordinates in its mouse
management, but it has the assumption that it is the top-level window
when restoring the mouse cursor position after dragging. Somehow it
will cause a crash in some cases. I've fixed that portion and I am not
able to crash it any more. (May be I am not trying hard enough 8-)

In windisplay.cpp

LRESULT
display::on_mouse( WPARAM wParam, LPARAM lParam)
...
if (mouse.is_mouse_locked() && ( mouse_x < 20 || mouse_x > view_width
- 20 || mouse_y < 20 || mouse_y > view_height - 20 ) ) {
		mouse.report_setcursor( view_width/2, view_height/2 );
		SetCursorPos( view_x + view_width/2, view_y + view_height/2 );
	}
	if (was_locked && !mouse.is_mouse_locked()) {
		mouse.report_setcursor( old_x, old_y );
		SetCursorPos( view_x + old_x, view_y + old_y );
	}
...


Both SetCursorPos call will move the mouse to an incorrect coordinate.
It is because WM_MOVE message is not sending to vPython any more after
it is reparented. (also, may be window_x and window_y shall be used
instead of view_x, view_y).

Changing it to:

	if (mouse.is_mouse_locked() && ( mouse_x < 20 || mouse_x > view_width
- 20 || mouse_y < 20 || mouse_y > view_height - 20 ) ) {
		mouse.report_setcursor( view_width/2, view_height/2 );
		//SetCursorPos( view_x + view_width/2, view_y + view_height/2 );
		POINT pt;
		pt.x = pt.y = 0;
		ClientToScreen(widget_handle, &pt);
		SetCursorPos( pt.x + view_width/2, pt.y + view_height/2 );
	}
	if (was_locked && !mouse.is_mouse_locked()) {
		mouse.report_setcursor( old_x, old_y );
		//SetCursorPos( view_x + old_x, view_y + old_y );
		POINT pt;
		pt.x = pt.y = 0;
		ClientToScreen(widget_handle, &pt);
		SetCursorPos( pt.x + old_x,  pt.y + old_y );
	}

Now the mouse cursor is correctly positioned after dragging, no matter
if it has been reparented or not. Seems no more crashes now.

CL

> >
> Looks something like VPython 5 is using absolute mouse coordinates instead
> of translating them to the position in the parent.
> > I tried SetWindowPos but it looks like it has no effect after the
> > scene window is created, so I change it to MoveWindow. It works very
> > well on my PC.
> >
> Now try this:
> - start the program
> - move the window more to the top of your screen
> - click and drag, from a bottom position in the VPython window, downwards
> in such a way that you virtual (you can't see the cursor) will leave the
> window,
> and I expect you get the following error:
> VPython ***CRITICAL ERROR***: ..\src\core\display_kernel.cpp:535:
> cvisual::display_kernel::world_to_view_transform: VPython degenerate
> projection: 1.#INF 1.#INF 0.57735 0.494872
>
> cheers,
> Stef
>
>
>
> Het UMC St Radboud staat geregistreerd bij de Kamer van Koophandel in het handelsregister onder nummer 41055629.
> The Radboud University Nijmegen Medical Centre is listed in the Commercial Register of the Chamber of Commerce under file number 41055629.

Re: [Visualpython-users] vpython and wxpython integration

[Stef M. <s.m...@ru...>]

hi Che ?

CL wrote:
> Hi Shef
>
> Since I only have one PC, I am not sure if my sample programs working
> well or not on other hardware. Are you able to run my sample programs
> on your side (or it also crashes) ? Anyone else can report the result
> if you have tried it ?
>   
aha, so you didn't try hard enough ;-)
> About the crashes, can you show a bit more about the python code to do
> SetWindowPos / MoveWindow ? Technically speaking, I can see no reason
> why using these API will cause a crash in general.
>
>   
Looks something like VPython 5 is using absolute mouse coordinates instead
of translating them to the position in the parent.
> I tried SetWindowPos but it looks like it has no effect after the
> scene window is created, so I change it to MoveWindow. It works very
> well on my PC.
>   
Now try this:
- start the program
- move the window more to the top of your screen
- click and drag, from a bottom position in the VPython window, downwards
in such a way that you virtual (you can't see the cursor) will leave the 
window,
and I expect you get the following error:
VPython ***CRITICAL ERROR***: ..\src\core\display_kernel.cpp:535: 
cvisual::display_kernel::world_to_view_transform: VPython degenerate 
projection: 1.#INF 1.#INF 0.57735 0.494872

cheers,
Stef



Het UMC St Radboud staat geregistreerd bij de Kamer van Koophandel in het handelsregister onder nummer 41055629.
The Radboud University Nijmegen Medical Centre is listed in the Commercial Register of the Chamber of Commerce under file number 41055629.

Re: [Visualpython-users] vpython and wxpython integration

[Stef M. <s.m...@ru...>]

hi Bruce,

Bruce Sherwood wrote:
> No, 
You might be completely right,
but the final result from my viewpoint is the same: "I can't use it for 
this purpose for a long period of time"
> the issue there was that there isn't any way for me to debug that very 
> complex program, since I know nothing about wxpython.
It has nothing to do wxPython.
And as I suspected that you didn't use wxPython,
I attached a little autoit-executable (which is written directly on the 
windows APIs), so you reproduce the bug without wxPython.
The fact that this autoit program shows the same problem,
also proves that it's not wxPython to blame.

cheers,
Stef



Het UMC St Radboud staat geregistreerd bij de Kamer van Koophandel in het handelsregister onder nummer 41055629.
The Radboud University Nijmegen Medical Centre is listed in the Commercial Register of the Chamber of Commerce under file number 41055629.

Re: [Visualpython-users] vpython and wxpython integration

[CL <clc...@gm...>]

Hi Shef

Since I only have one PC, I am not sure if my sample programs working
well or not on other hardware. Are you able to run my sample programs
on your side (or it also crashes) ? Anyone else can report the result
if you have tried it ?

About the crashes, can you show a bit more about the python code to do
SetWindowPos / MoveWindow ? Technically speaking, I can see no reason
why using these API will cause a crash in general.

I tried SetWindowPos but it looks like it has no effect after the
scene window is created, so I change it to MoveWindow. It works very
well on my PC.

best regards,
CL
> I just tried MoveWindow instead of SetWindowPos, but it crashes just as
> hard with VPython-5.
> Here is the essential part of the code:
>
>        # get the handle of the dock container
>        PP = self.p1.GetHandle ()
>
>        # Set Position and Size of the VPython window,
>        # Before Docking it !!
>        flags = win32con.SWP_SHOWWINDOW or \
>                win32con.SWP_FRAMECHANGED
>        win32gui.SetWindowPos ( self.VP, win32con.HWND_TOPMOST,
>                                -4, -22, w+8, h+26, flags )
>        #win32gui.MoveWindow ( self.VP, -4, -22, w+8, h+26, True )
> <===JUST AS BAD
>
>        # Dock the VPython window
>        win32gui.SetParent ( self.VP, PP )
>
> Am I doing something wrong,
> or are you not testing hard enough ?
>
> cheers,
> Stef
>
> > This examples only work on Windows platform because the docking
> > technique is depending on win32 api. You can change it to non-docking
> > version easily by removing some code in vWorld object.
> >
> > Example 1 is just the bounce2 sample program implemented in a wxdialog box.
> > Example 2 is four bounce2 windows in a wxdialog box.
> >
> > Hope this can help you.
> >
> > ------------------------------------------------------------------------------
> > This SF.net email is sponsored by:
> > SourcForge Community
> > SourceForge wants to tell your story.
> > http://p.sf.net/sfu/sf-spreadtheword
> > _______________________________________________
> > Visualpython-users mailing list
> > Vis...@li...
> > https://lists.sourceforge.net/lists/listinfo/visualpython-users
> >
> >
> >
>
>
>
> ------------------------------
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf-spreadtheword
>
> ------------------------------
>
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users
>
>
> End of Visualpython-users Digest, Vol 32, Issue 11
> **************************************************
>

Re: [Visualpython-users] Opacity problem

[CL <clc...@gm...>]

Hi Bruce

Not at all. Unfortunately I am not very experienced in low level
graphics/opengl development, I can only help at surface level / minor
logic bug level.

CL

On 1/15/09, Bruce Sherwood <Bru...@nc...> wrote:
> Many thanks for looking into this! I'll try it.
>
> More generally, it would be a much healthier situation for VPython if more
> users would/could get involved at this level. Such help is greatly
> appreciated by everyone in the community.
>
> Bruce Sherwood
>
> CL wrote:
> >
> > About the opacity problem (crash when changing from opacity 1.0 to
> > lower), it looks like it is a bug in frame.cpp
> >
> > void frame::gl_render( const view& v)
> >
> > ....
> >
> > for (child_iterator i = children.begin(); i !=
> > child_iterator(children.end()); ++i) {
> >        if (i->translucent()) {
> >                // See display_kernel::draw().
> >                trans_children.push_back( *i.base());
> >                i = children.erase(i.base());
> >                continue;
> >        }
> >
> >        i->outer_render(local);
> > }
> >
> > The code is trying to move a renderable from a list to another list
> > when the opacity of the renderable is changing from 1.0 to something
> > else. But children.erase() function returns the pointer after
> > deletion, not the pointer before deletion.
> >
> > Following similar logic in display_kernel::draw, replace the code with :
> >
> > child_iterator i(children.begin());
> > child_iterator i_end(children.end());
> > while (i != i_end) {
> >        if (i->translucent()) {
> >                // See display_kernel::draw().
> >                trans_children.push_back( *i.base());
> >                i = children.erase(i.base());
> >                continue;
> >        }
> >        i->outer_render(local);
> >        i++;
> >
> > It seems working on my simple test cases.
> >
> >
> >
> > > Don't I wish it were that simple. No, objects are created with
> > > opacity=1.0 by default. That's why a scene contains opaque objects if
> > > you don't explicitly set the opacity to something different.
> > >
> >
> >
> > > b = box()
> > > print b.opacity # shows 1.0
> > >
> >
> >
> > > Bruce Sherwood
> > >
> >
> >
> ------------------------------------------------------------------------------
> > This SF.net email is sponsored by:
> > SourcForge Community
> > SourceForge wants to tell your story.
> > http://p.sf.net/sfu/sf-spreadtheword
> > _______________________________________________
> > Visualpython-users mailing list
> > Vis...@li...
> >
> https://lists.sourceforge.net/lists/listinfo/visualpython-users
> >
>

Re: [Visualpython-users] vpython and wxpython integration

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

No, the issue there was that there isn't any way for me to debug that 
very complex program, since I know nothing about wxpython. In the case 
of Mac Tk, the issue is known: Carbon wants to be the primary thread. I 
don't know what property of Visual 5 interferes with trying to embed it 
in wxpython on Windows.

Bruce Sherwood

Stef Mientki wrote:
> hi Bruce,
> 
> Bruce Sherwood wrote:
>> I know about the limitation of Visual 5 that on the Mac it's not 
>> possible to invoke Tk, due to conflicts over who is the primary 
>> thread, but what's the nature of the issue with wxpython on Windows? I 
>> don't recall saying anything about features of Visual 5 that would 
>> prevent doing something with wxpython. (I wouldn't be surprised if 
>> there were problems, given the architecture of Visual, it's just that 
>> I don't know anything about wxpython, nor what aspects of Visual 5 
>> would cause problems.)
> Well you gave the answer (in which the "features" were mentioned)  off list,
> the problem was docking VPython-5, after moving or resizing the window,
> VPython crashed.
> 
> cheers,
> Stef
> 
> 
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf-spreadtheword
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users

Re: [Visualpython-users] vpython and wxpython integration

[Stef M. <s.m...@ru...>]

Emanuele Gissi wrote:
> Thanks Stef for your reply, but my application has to be multiplatform.
> I think that your solution is for Windows only.
>
> I don't need wxpython in particular, but the vpython controls are not 
> enough for me.
Now the Qt license will change,
it might be interesting to take a look at Qt,
seems to have a dock feature for windows and linux ( don't know for MAC).

cheers,
Stef

Re: [Visualpython-users] vpython and wxpython integration

[Stef M. <s.m...@ru...>]

hi Bruce,

Bruce Sherwood wrote:
> I know about the limitation of Visual 5 that on the Mac it's not 
> possible to invoke Tk, due to conflicts over who is the primary 
> thread, but what's the nature of the issue with wxpython on Windows? I 
> don't recall saying anything about features of Visual 5 that would 
> prevent doing something with wxpython. (I wouldn't be surprised if 
> there were problems, given the architecture of Visual, it's just that 
> I don't know anything about wxpython, nor what aspects of Visual 5 
> would cause problems.)
Well you gave the answer (in which the "features" were mentioned)  off list,
the problem was docking VPython-5, after moving or resizing the window,
VPython crashed.

cheers,
Stef

Re: [Visualpython-users] vpython and wxpython integration

[Stef M. <s.m...@ru...>]

CL wrote:
> It seems there are quite a number of developers looking for simple
> examples showing an integration of wxpython and vpython, I have just
> written two examples and put it here:
> http://code.google.com/p/vpythonex1/  (as my robot simulator is quite
> complicated).
>
> In these examples, I try to use the docking technique described in
> http://mientki.ruhosting.nl/data_www/pylab_works/pw_vpython_docking.html
>
> I use MoveWindow instead of SetWindowPos, in order to make it work on
> the new version of vPython. I have not tested them on previous verion
> of vPython.
>
>   
I just tried MoveWindow instead of SetWindowPos, but it crashes just as 
hard with VPython-5.
Here is the essential part of the code:

        # get the handle of the dock container
        PP = self.p1.GetHandle ()

        # Set Position and Size of the VPython window,
        # Before Docking it !!
        flags = win32con.SWP_SHOWWINDOW or \
                win32con.SWP_FRAMECHANGED
        win32gui.SetWindowPos ( self.VP, win32con.HWND_TOPMOST,
                                -4, -22, w+8, h+26, flags )
        #win32gui.MoveWindow ( self.VP, -4, -22, w+8, h+26, True )   
<===JUST AS BAD
       
        # Dock the VPython window
        win32gui.SetParent ( self.VP, PP )

Am I doing something wrong,
or are you not testing hard enough ?

cheers,
Stef

> This examples only work on Windows platform because the docking
> technique is depending on win32 api. You can change it to non-docking
> version easily by removing some code in vWorld object.
>
> Example 1 is just the bounce2 sample program implemented in a wxdialog box.
> Example 2 is four bounce2 windows in a wxdialog box.
>
> Hope this can help you.
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf-spreadtheword
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users
>
>
>   

Re: [Visualpython-users] vpython and wxpython integration

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

This is probably a naive notion, but this sounds like a somewhat 
straight-forward VPython program, in which you show a tree of the 
displayed objects (probably the nodes of the tree are miniature versions 
of those objects). I assume by tree you mean that objects would be 
grouped under frames. I have the vague notion that someone once built 
some kind of visual tree structure in VPython....? Or that there exist 
Python modules for creating/analyzing tree structures?

I also draw everyone's attention to the new component of Visual 5, 
filedialog.py, which is an example of pure platform-independent Python 
code which provides basic file dialog services. It also works with 
Visual 3. Maybe there are elements there of the additional controls you 
seek.

Bruce Sherwood

Emanuele Gissi wrote:
> Thanks Stef for your reply, but my application has to be multiplatform.
> I think that your solution is for Windows only.
> 
> I don't need wxpython in particular, but the vpython controls are not 
> enough for me.
> I would like to show a tree of the displayed objects, so that the user 
> can select them and edit their properties. This could be in another 
> window, thought I would prefer it integrated in the 3dview window.
> 
> Can someone suggest a tool to obtain such an effect?
> Thanks once more.
> Emanuele Gissi
> 
> 
> ------------------------------------------------------------------------
> 
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf-spreadtheword
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users

Re: [Visualpython-users] Opacity problem

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

Many thanks for looking into this! I'll try it.

More generally, it would be a much healthier situation for VPython if 
more users would/could get involved at this level. Such help is greatly 
appreciated by everyone in the community.

Bruce Sherwood

CL wrote:
> About the opacity problem (crash when changing from opacity 1.0 to
> lower), it looks like it is a bug in frame.cpp
> 
> void frame::gl_render( const view& v)
> 
> ....
> 
> for (child_iterator i = children.begin(); i !=
> child_iterator(children.end()); ++i) {
> 	if (i->translucent()) {
> 		// See display_kernel::draw().
> 		trans_children.push_back( *i.base());
> 		i = children.erase(i.base());
> 		continue;
> 	}
> 
> 	i->outer_render(local);
> }
> 
> The code is trying to move a renderable from a list to another list
> when the opacity of the renderable is changing from 1.0 to something
> else. But children.erase() function returns the pointer after
> deletion, not the pointer before deletion.
> 
> Following similar logic in display_kernel::draw, replace the code with :
> 
> child_iterator i(children.begin());
> child_iterator i_end(children.end());
> while (i != i_end) {
> 	if (i->translucent()) {
> 		// See display_kernel::draw().
> 		trans_children.push_back( *i.base());
> 		i = children.erase(i.base());
> 		continue;
> 	}
> 	i->outer_render(local);
> 	i++;
> 
> It seems working on my simple test cases.
> 
> 
>> Don't I wish it were that simple. No, objects are created with
>> opacity=1.0 by default. That's why a scene contains opaque objects if
>> you don't explicitly set the opacity to something different.
> 
>> b = box()
>> print b.opacity # shows 1.0
> 
>> Bruce Sherwood
> 
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf-spreadtheword
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users

Re: [Visualpython-users] Difficulty in compilation on Ubuntu 8.10 x86_64

[Suman C. <su...@ss...>]

Bruce Sherwood wrote:
> You might not have to compile it. Isn't there already a Ubuntu package
> "python-visual"? Or is the issue that your machine is 64 bit?


Thank you very much! I had no idea about the package name, so first I
tried the compilation as I do with any other new package.

I have installed python-visual successfully, but of course that is
version 3.2.9. There might still be some problem with the version 5. But
in any case it will serve my purpose for the time being.


Thanks and regards,
Suman Chakrabarty.



-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.

[Visualpython-users] vpython and wxpython integration

[Emanuele G. <ema...@gm...>]

Thanks Stef for your reply, but my application has to be multiplatform.
I think that your solution is for Windows only.

I don't need wxpython in particular, but the vpython controls are not enough
for me.
I would like to show a tree of the displayed objects, so that the user can
select them and edit their properties. This could be in another window,
thought I would prefer it integrated in the 3dview window.

Can someone suggest a tool to obtain such an effect?
Thanks once more.
Emanuele Gissi

[Visualpython-users] Opacity problem

[CL <clc...@gm...>]

About the opacity problem (crash when changing from opacity 1.0 to
lower), it looks like it is a bug in frame.cpp

void frame::gl_render( const view& v)

....

for (child_iterator i = children.begin(); i !=
child_iterator(children.end()); ++i) {
	if (i->translucent()) {
		// See display_kernel::draw().
		trans_children.push_back( *i.base());
		i = children.erase(i.base());
		continue;
	}

	i->outer_render(local);
}

The code is trying to move a renderable from a list to another list
when the opacity of the renderable is changing from 1.0 to something
else. But children.erase() function returns the pointer after
deletion, not the pointer before deletion.

Following similar logic in display_kernel::draw, replace the code with :

child_iterator i(children.begin());
child_iterator i_end(children.end());
while (i != i_end) {
	if (i->translucent()) {
		// See display_kernel::draw().
		trans_children.push_back( *i.base());
		i = children.erase(i.base());
		continue;
	}
	i->outer_render(local);
	i++;

It seems working on my simple test cases.


>Don't I wish it were that simple. No, objects are created with
>opacity=1.0 by default. That's why a scene contains opaque objects if
>you don't explicitly set the opacity to something different.

>b = box()
>print b.opacity # shows 1.0

>Bruce Sherwood

Re: [Visualpython-users] vpython and wxpython integration

[CL <clc...@gm...>]

It seems there are quite a number of developers looking for simple
examples showing an integration of wxpython and vpython, I have just
written two examples and put it here:
http://code.google.com/p/vpythonex1/  (as my robot simulator is quite
complicated).

In these examples, I try to use the docking technique described in
http://mientki.ruhosting.nl/data_www/pylab_works/pw_vpython_docking.html

I use MoveWindow instead of SetWindowPos, in order to make it work on
the new version of vPython. I have not tested them on previous verion
of vPython.

This examples only work on Windows platform because the docking
technique is depending on win32 api. You can change it to non-docking
version easily by removing some code in vWorld object.

Example 1 is just the bounce2 sample program implemented in a wxdialog box.
Example 2 is four bounce2 windows in a wxdialog box.

Hope this can help you.

Re: [Visualpython-users] vpython and wxpython integration

[CL <clc...@gm...>]

If you are looking for integration with separated windows (not
embedding vpython into a child window in wxpython, you can see if this
can help:
http://code.google.com/p/clrobotsim/

Basically you can just create a separated thread for the vpython
object, use standard communication methods between wxpython and
vpython.
Dead lock may happen in the wxpython GUI thread if you do not handle
it properly. You can check the code at the "robot executor" of the
above simulator.

> From: "Emanuele Gissi" <ema...@gm...>
> Subject: [Visualpython-users] vpython and wxpython integration
> To: vis...@li...
> Message-ID:
>        <e00...@ma...>
> Content-Type: text/plain; charset="iso-8859-1"
>
> I am developing a small multiplatform application for data visualisation.
> Could someone point me to a simple example on how to integrate a wxpython
> GUI with vpython 3d visualisation?
> Thank you.
> Emanuele Gissi

Re: [Visualpython-users] Difficulty in compilation on Ubuntu 8.10 x86_64

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

Thanks for reporting this!

Bruce Sherwood

David Roundy wrote:
> On Wed, Jan 14, 2009 at 11:57 AM, Bruce Sherwood wrote:
>> You might not have to compile it. Isn't there already a Ubuntu package
>> "python-visual"? Or is the issue that your machine is 64 bit?
> 
> I can confirm that the python-visual package for ubuntu works fine on
> 64 bit machines, for what it's worth...
> 
> David

Re: [Visualpython-users] Difficulty in compilation on Ubuntu 8.10 x86_64

[David R. <ro...@ph...>]

On Wed, Jan 14, 2009 at 11:57 AM, Bruce Sherwood wrote:
> You might not have to compile it. Isn't there already a Ubuntu package
> "python-visual"? Or is the issue that your machine is 64 bit?

I can confirm that the python-visual package for ubuntu works fine on
64 bit machines, for what it's worth...

David

Re: [Visualpython-users] vpython and wxpython integration

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

I know about the limitation of Visual 5 that on the Mac it's not 
possible to invoke Tk, due to conflicts over who is the primary thread, 
but what's the nature of the issue with wxpython on Windows? I don't 
recall saying anything about features of Visual 5 that would prevent 
doing something with wxpython. (I wouldn't be surprised if there were 
problems, given the architecture of Visual, it's just that I don't know 
anything about wxpython, nor what aspects of Visual 5 would cause problems.)

Bruce Sherwood

Stef Mientki wrote:
> 
> Emanuele Gissi wrote:
>> I am developing a small multiplatform application for data visualisation.
>> Could someone point me to a simple example on how to integrate a 
>> wxpython GUI with vpython 3d visualisation?
> look here,
> 
> http://mientki.ruhosting.nl/data_www/pylab_works/pw_vpython_docking.html
> 
> but it will only work for VPython-3.
> In VPython-5 there are "features" that will crash the application,
> and from Bruce I understood that these "features"  will not be removed 
> on a short term.
> 
> cheers,
> Stef
> 
>> Thank you.
>> Emanuele Gissi
>> ------------------------------------------------------------------------
>>
>> ------------------------------------------------------------------------------
>> This SF.net email is sponsored by:
>> SourcForge Community
>> SourceForge wants to tell your story.
>> http://p.sf.net/sfu/sf-spreadtheword
>> ------------------------------------------------------------------------
>>
>> _______________________________________________
>> Visualpython-users mailing list
>> Vis...@li...
>> https://lists.sourceforge.net/lists/listinfo/visualpython-users
>>   
> 
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf-spreadtheword
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users

Re: [Visualpython-users] Difficulty in compilation on Ubuntu 8.10 x86_64

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

The issue is that there has been very little experience of people 
building Visual with 64-bit Linux. I myself don't have access to a 
64-bit machine. Looking at a build on 32-bit Ubuntu, I don't see 
anything in your build.log or config.log that explains why there's some 
incompatibility with the includes.

I do however see something that makes me wonder. In config.log at the 
start of the section labeled "Cache variables" you have

ac_cv_build=x86_64-unknown-linux-gnu
ac_cv_build_alias=x86_64-unknown-linux-gnu

while I see this:

ac_cv_build=i686-pc-linux-gnu

Makes me wonder about the "unknown" in your listing, though maybe that's 
irrelevant.

One other thing: What I have on my Ubuntu machine is 
libboost-python1.34.1, which is the only version I see in the synaptic 
package manager.

All I can hope is that someone else on this list has experience with 
64-bit Ubuntu and can offer some help.

Bruce Sherwood

Suman Chakrabarty wrote:
> Dear all,
> 
> I am trying to compile visual-3.2.9 on a 64 bit Ubuntu 8.10 and getting
> the following problem at the "make" step:
> 
> =============================================================================
> <snip>
> Updating dependancy information for ../../visual-3.2.9/src/arrow.cpp ...
> make[1]: Leaving directory `/home/suman/Softwares/VPython/src'
> make[1]: Entering directory `/home/suman/Softwares/VPython/src'
> This is a quiet Makefile. If make exits with an error, check
> src/build.log to see the complete error message(s). In the event of an
> error that you cannot debug, please send a message to
> vis...@li..., including the files config.log
> and src/build.log, requesting assistance.
> Compiling ../../visual-3.2.9/src/arrow.cpp ...
> make[1]: *** [arrow.lo] Error 1
> make[1]: Leaving directory `/home/suman/Softwares/VPython/src'
> make: *** [all-recursive] Error 1
> ==============================================================================
> 
> Last two lines containing error messages in build.log:
> 
> ../../visual-3.2.9/include/platlinux.h:27: error: declaration of
> 'typedef class visual::lock<visual::mutex> visual::mutex::lock'
> ../../visual-3.2.9/include/vthread.h:13: error: changes meaning of
> 'lock' from 'class visual::lock<visual::mutex>'
> ===============================================================================
> 
> Since I have no clue about how to debug this problem, I have attached
> the config.log and build.log files. It will be much appreciated if
> somebody can help me get through this.
> 
> 
> Thanks and regards,
> Suman Chakrabarty.
> 
> 
> 
> 
> 
> 
> ------------------------------------------------------------------------
> 
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf-spreadtheword
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users

Re: [Visualpython-users] Difficulty in compilation on Ubuntu 8.10 x86_64

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

You might not have to compile it. Isn't there already a Ubuntu package 
"python-visual"? Or is the issue that your machine is 64 bit?

Bruce Sherwood

Suman Chakrabarty wrote:
> Dear all,
> 
> I am trying to compile visual-3.2.9 on a 64 bit Ubuntu 8.10 and getting
> the following problem at the "make" step:
> 
> =============================================================================
> <snip>
> Updating dependancy information for ../../visual-3.2.9/src/arrow.cpp ...
> make[1]: Leaving directory `/home/suman/Softwares/VPython/src'
> make[1]: Entering directory `/home/suman/Softwares/VPython/src'
> This is a quiet Makefile. If make exits with an error, check
> src/build.log to see the complete error message(s). In the event of an
> error that you cannot debug, please send a message to
> vis...@li..., including the files config.log
> and src/build.log, requesting assistance.
> Compiling ../../visual-3.2.9/src/arrow.cpp ...
> make[1]: *** [arrow.lo] Error 1
> make[1]: Leaving directory `/home/suman/Softwares/VPython/src'
> make: *** [all-recursive] Error 1
> ==============================================================================
> 
> Last two lines containing error messages in build.log:
> 
> ../../visual-3.2.9/include/platlinux.h:27: error: declaration of
> 'typedef class visual::lock<visual::mutex> visual::mutex::lock'
> ../../visual-3.2.9/include/vthread.h:13: error: changes meaning of
> 'lock' from 'class visual::lock<visual::mutex>'
> ===============================================================================
> 
> Since I have no clue about how to debug this problem, I have attached
> the config.log and build.log files. It will be much appreciated if
> somebody can help me get through this.
> 
> 
> Thanks and regards,
> Suman Chakrabarty.
> 
> 
> 
> 
> 
> 
> ------------------------------------------------------------------------
> 
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf-spreadtheword
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users