visualpython-users

[Visualpython-users] (no subject)

[Jason M. <j.a...@gm...>]

Hi,

 I've tried to build from source on Ubuntu 10.10, but ran into the following error during link.

 /bin/sed: can't read /usr/lib/libgdk_pixbuf-2.0.la: No such file or directory
 libtool: link: `/usr/lib/libgdk_pixbuf-2.0.la' is not a valid libtool archive

 It appears that is related to this:
http://www.debian.org/doc/debian-policy/ch-files.html#s-libraries 
 referred to by this thread:
https://bugs.launchpad.net/ubuntu/+source/gdk-pixbuf/+bug/665768 
 Where, about half way down it says that .la files are not longer distributed if they don't conform.

 There is a workround (which I've not yet had time to try).

 But perhaps the proper fix should be applied?

 Cheers,
 Jason.

Re: [Visualpython-users] VPython mac os

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

scene.x and scene.y control the placement of the window. What you want
are scene.width and scene.height.

Bruce Sherwood

On Thu, Mar 31, 2011 at 2:59 AM, Georges Arsouze
<geo...@gm...> wrote:
> Hi
> Sorry for my poor English
> I'm working with visual python under python 3 under mac os 10.6.7
> I'm just beginning
> I don't understand why scene.width and scene.height don't woks on my
> computer
> (If I change the values in the program nothing change)
> from visual import *
> scene.x=800
> scene.y=120
> scene.title='Thompson1'
> scene.widh=800
> scene.height=400
> #scene.autoscale=0
> #scene.range=(100,100,100)
> #scene.center=(0,100,0)
> balle=sphere(pos=(0,80,0),radius=2)
> sol=box(pos=(0,-1,0), size=(10,2,10))
> Regards
> ------------------------------------------------------------------------------
> Create and publish websites with WebMatrix
> Use the most popular FREE web apps or write code yourself;
> WebMatrix provides all the features you need to develop and
> publish your website. http://p.sf.net/sfu/ms-webmatrix-sf
>
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users
>
>

[Visualpython-users] VPython mac os

[Georges A. <geo...@gm...>]

Hi
Sorry for my poor English

I'm working with visual python under python 3 under mac os 10.6.7
I'm just beginning
I don't understand why scene.width and scene.height don't woks on my
computer
(If I change the values in the program nothing change)

from visual import *
scene.x=800
scene.y=120

scene.title='Thompson1'

scene.widh=800
scene.height=400
#scene.autoscale=0
#scene.range=(100,100,100)
#scene.center=(0,100,0)

balle=sphere(pos=(0,80,0),radius=2)
sol=box(pos=(0,-1,0), size=(10,2,10))

Regards

[Visualpython-users] VPython 5.70: easy way to add a trail to a moving object

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

VPython 5.70 now available at vpython.org adds an easy to add a trail
to a moving object:

ball = sphere(.... make_trail=True ....)

The object can be an arrow, box, cone, cylinder, ellipsoid, pyramid,
ring, or sphere

If you create the object with make_trail=False, the trail won't start
until and unless you later say make_trail=True. In order to make any
trail you must specify make_trail to be True of False when you create
the object.

Additional options:

trail_type     Default is "curve", but can be "points"

interval        If interval=10, a point is added to the trail only
every 10th move of the object

retain          If retain=50, only the 50 most recently added points
will be retained

trail_object  The curve or points object;
ball.trail_object.color=color.orange will change the color of the
trail to orange

There were two motivations for offering this. The first is simply that
leaving a trail behind a moving object is a common task in VPython,
and it is a bit annoying to have to create a curve object and do
appends with or without retains. It is a bit more annoying to specify
skipping points.

A second major reason for implementing easier trail creation comes
from recent research by Shawn Weatherford at North Carolina State
University. Shawn, a Ph.D. student of Ruth Chabay, ran a lab section
in which students gave permission for extensive video and audio
recording of their work throughout the semester, which included
experiments, group problem-solving, and computational modeling using
VPython. Shawn presented his data last week and showed that one of the
most common student difficulties had to do with making an object leave
a trail.

These are first-year and second-year engineering and science students
in an introductory physics course. The students are facile with
computers except for programming, which is new to almost all of them.
Just as Stephen Hawking's publisher warned him that every equation he
included in "A Brief History of Time" would cost him half his
potential readership, so in this programming environment every
non-physics element can cause a significant rise in student
difficulty, because it distracts attention away from concentrating on
the physics and toward language syntax.

Particularly alarming was the observation that when students were
asked to read through a partially complete program, to exercise the
important skill of being able to read a program, frequently the
students would see "curve" (needed to implement a trail) and say, "Oh,
the spaceship is going to curve, and here is the programming statement
that will make it curve"!!!! Sigh. Hopefully adding "make_trail=True"
won't trigger this kind of mindlessness.

Bruce Sherwood

Re: [Visualpython-users] Camera panning (or something like it)

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

A pan mode would indeed be useful, and as you saw in an obsolete
"4beta" page it briefly existed. At about that time there was a big
upheaval in VPython development, and in the process of major progress
it wasn't possible to keep the beta-level pan mechanism. It is
certainly something worth implementing in the future.

I draw your attention to an entry in the "contributed programs"
section of vpython.org:

    Zoom and spin under complete program control (Bruce Sherwood)

You could probably enhance this with not much effort to offer a pan mode.

If your graphics card doesn't have an acceleration mode, I don't see
how you can make the hardware go faster.

Bruce Sherwood

On Sat, Mar 26, 2011 at 9:27 AM, Christian Jauvin <cj...@gm...> wrote:
> Hi,
>
> Please excuse in advance my "3d newbiness", as I might be missing some
> obvious background knowledge, or my question not be totally clear.
>
> I am rendering a complex point cloud scene and while the default zoom
> and spin mouse interface is working well, I find that I miss the
> ability to pan the camera (or the scene) to obtain a better view.
>
> I have played with what I found in many example programs (i.e.
> modifying the scene.center vector in a simple way), but it's never
> exactly what I need. The only reference I found to what I (possibly)
> need is on this page:
>
> http://vpython.org/webdoc/texture-opacity-lighting.html
>
> where they mention a new "pan" mode. But then it doesn't seem to be
> implemented in the version I'm running, which is 5.12 on Ubuntu 10.04
> 64-bit.
>
> And finally, I'd like to know if there are some settings that could
> enhance VPython's performance on my non-3d-accelerated laptop.
>
> Best regards,
>
> Christian
>
> ------------------------------------------------------------------------------
> Enable your software for Intel(R) Active Management Technology to meet the
> growing manageability and security demands of your customers. Businesses
> are taking advantage of Intel(R) vPro (TM) technology - will your software
> be a part of the solution? Download the Intel(R) Manageability Checker
> today! http://p.sf.net/sfu/intel-dev2devmar
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users
>

[Visualpython-users] Camera panning (or something like it)

[Christian J. <cj...@gm...>]

Hi,

Please excuse in advance my "3d newbiness", as I might be missing some
obvious background knowledge, or my question not be totally clear.

I am rendering a complex point cloud scene and while the default zoom
and spin mouse interface is working well, I find that I miss the
ability to pan the camera (or the scene) to obtain a better view.

I have played with what I found in many example programs (i.e.
modifying the scene.center vector in a simple way), but it's never
exactly what I need. The only reference I found to what I (possibly)
need is on this page:

http://vpython.org/webdoc/texture-opacity-lighting.html

where they mention a new "pan" mode. But then it doesn't seem to be
implemented in the version I'm running, which is 5.12 on Ubuntu 10.04
64-bit.

And finally, I'd like to know if there are some settings that could
enhance VPython's performance on my non-3d-accelerated laptop.

Best regards,

Christian

[Visualpython-users] VPython 5.61: crosshairs in graphs, bug fixes

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

With the new VPython 5.61 available at vpython.org you can drag the
mouse across a graph and see crosshairs with the (x,y) value of the
crosshair position, to make it possible to read locations accurately
off the graph. No change is needed to existing programs to get this
behavior.

If you use the controls module (buttons, sliders, menus), you no
longer need to call interact() inside a loop. For backward
compatibility the interact() function is still present but now does
nothing.

A long-standing inconsistency in the label object has been fixed. The
background of a label object's box is now by default the same as
scene.background. You can also set the label's background color.

There are bug fixes to materials.bricks, thanks to Kadir Haldenbilen,
and a bug fix to the machinery that lets VPython run even if the
Polygon and font modules are unavailable (though in this case you
can't use extrusion or text objects).

Bruce Sherwood

P.S. In the course of this work I was surprised to discover what seems
to be a long-standing bug on the Mac. In programs that open two
graphics windows, occasionally (maybe 1% of the time) the second
window fails to open and, even more exciting, Python statements stop
being executed! The first window remains active, and you can zoom and
spin. Be on the lookout for such behavior on the Mac. It would be
great if there were some program that failed every time. As it is, the
failure happens so seldom that it is difficult to debug.

I first noticed this in the graphtest.py example program and thought I
must have broken something in the process of enabling crosshairs, but
then I found the problem was present in older versions, and in
programs that don't create graphing windows.

Re: [Visualpython-users] Problems while Installing 5.60 on Ubuntu 10.04

[Thomas S. <to...@fe...>]

-lGL is an argument, that tells the linker to search for the library
"libGL.so" and link against it.
According to [1] it looks like it's libgl1-mesa-dev or ia32-libs.

	Thomas

[1]
http://packages.ubuntu.com/search?searchon=contents&keywords=libGL.so&mode=exactfilename&suite=maverick&arch=any

On Thu, 24 Mar 2011 17:27:57 -0600
Bruce Sherwood wrote:

> I'm guessing that GL refers to the libgl1-mesa packages, which are an
> open-source version of OpenGL. I think that normally they're already
> installed, but apparently not on your machine. Anyone with better
> information is invited to comment!
> 
> Bruce
> 
> 2011/3/24 Stefan Hüller <ste...@if...>:
> > Hello Bruce and experts,
> >
> > I hope that it is o.k. for you when I send my config.log and
> > build.log in a zip-file as you mentioned in the INSTALL.TXT for
> > Visual 5.60! I tried hard to follow up your installation tutorial,
> > and I was happy until the build.log ended with the last two lines:
> > "/usr/bin/ld: cannot find -lGL
> > collect2: ld returned 1 exit status".
> >
> > Would you mind giving me a hint so I can follow up? All references I
> > googled couldn't satisfy.
> >
> > Thank you in advance,
> > Stefan

Re: [Visualpython-users] Problems while Installing 5.60 on Ubuntu 10.04

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

I'm guessing that GL refers to the libgl1-mesa packages, which are an
open-source version of OpenGL. I think that normally they're already
installed, but apparently not on your machine. Anyone with better
information is invited to comment!

Bruce

2011/3/24 Stefan Hüller <ste...@if...>:
> Hello Bruce and experts,
>
> I hope that it is o.k. for you when I send my config.log and build.log
> in a zip-file as you mentioned in the INSTALL.TXT for Visual 5.60!
> I tried hard to follow up your installation tutorial, and I was happy
> until the build.log ended with the last two lines:
> "/usr/bin/ld: cannot find -lGL
> collect2: ld returned 1 exit status".
>
> Would you mind giving me a hint so I can follow up? All references I
> googled couldn't satisfy.
>
> Thank you in advance,
> Stefan
>
>
> ------------------------------------------------------------------------------
> Enable your software for Intel(R) Active Management Technology to meet the
> growing manageability and security demands of your customers. Businesses
> are taking advantage of Intel(R) vPro (TM) technology - will your software
> be a part of the solution? Download the Intel(R) Manageability Checker
> today! http://p.sf.net/sfu/intel-dev2devmar
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users
>
>

[Visualpython-users] Problems while Installing 5.60 on Ubuntu 10.04

[Stefan H. <ste...@if...>]

Hello Bruce and experts,

I hope that it is o.k. for you when I send my config.log and build.log
in a zip-file as you mentioned in the INSTALL.TXT for Visual 5.60!
I tried hard to follow up your installation tutorial, and I was happy
until the build.log ended with the last two lines:
"/usr/bin/ld: cannot find -lGL
collect2: ld returned 1 exit status".

Would you mind giving me a hint so I can follow up? All references I
googled couldn't satisfy.

Thank you in advance,
Stefan

Re: [Visualpython-users] Raytracing VPython

[K.-Michael A. <kmi...@gm...>]

On 2011-03-24 02:04:11 +0100, Anton Sherwood said:

> On 2011-3-23 16:55, Bruce Sherwood wrote:
>> But I don't think POVray will provide quantitative information about
>> the illumination at various places on the height field, will it?
> 
> One might say that's the primary thing that Povray does provide (in such
> a simple case).
> 
> I don't remember whether Povray ordinarily takes into account the
> distance between the light source and the object.  It's not clear from
> the inquiry whether that's desired here.

I can calculate the distance to the sun myself, just need to know which 
points get light when.
So I guess it's finally time to learn Povray *gulp*...


> 
> It would seem simplest to skip the VP step and write a Povray
> description of the height-field.  Use an orthographic camera!

Thanks all for your input!

Best regards,
Michael

Re: [Visualpython-users] Raytracing VPython

[Anton S. <br...@po...>]

On 2011-3-23 16:55, Bruce Sherwood wrote:
> But I don't think POVray will provide quantitative information about
> the illumination at various places on the height field, will it?

One might say that's the primary thing that Povray does provide (in such 
a simple case).

I don't remember whether Povray ordinarily takes into account the 
distance between the light source and the object.  It's not clear from 
the inquiry whether that's desired here.

It would seem simplest to skip the VP step and write a Povray 
description of the height-field.  Use an orthographic camera!

-- 
Anton Sherwood *\\* www.bendwavy.org *\\* www.zazzle.com/tamfang

Re: [Visualpython-users] Raytracing VPython

[Daπid <dav...@gm...>]

I have been doing some research and I think it is possible to export
to Blender somehow. I will try my minions to work on that, or do it
myself.

Anyway, it is a long-term project.

On Thu, Mar 24, 2011 at 1:20 AM, C Anthony Risinger <an...@ex...> wrote:
> On Mar 23, 2011 1:06 AM, "K. -Michael Aye" <mic...@sp...>
> wrote:
>>
>> Dear VPython folks,
>>
>> I was wondering if one can use VPython to do the following:
>>
>> * take a digital elevation map (DEM) and load it into a numpy array
>> * place a light source at some coordinate
>> * check for each point on the DEM if they can see the light source or if
>> they are being shadowed by other points of the DEM?
>>
>> As some kind of lightning calculations are done in the beautiful sceneries
>> of VPython, I wondered if I can 'abuse' it this way somehow?
>>
>> In case it can be done, I would need to know though, how accurate this
>> would be, as I would like to calculate solar irradiation times on the DEM
>> coordinates (just to make you interested! ;)
>>
>> Any suggestions welcome!
>
> About 5 years ago I used `microdem` to estimate radio coverage of pre-wimax
> equipment ... maybe its useful in your case.
>
> C Anthony [mobile]
>
> ------------------------------------------------------------------------------
> Enable your software for Intel(R) Active Management Technology to meet the
> growing manageability and security demands of your customers. Businesses
> are taking advantage of Intel(R) vPro (TM) technology - will your software
> be a part of the solution? Download the Intel(R) Manageability Checker
> today! http://p.sf.net/sfu/intel-dev2devmar
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users
>
>

Re: [Visualpython-users] Raytracing VPython

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

On Mar 23, 2011 1:06 AM, "K. -Michael Aye" <mic...@sp...>
wrote:
>
> Dear VPython folks,
>
> I was wondering if one can use VPython to do the following:
>
> * take a digital elevation map (DEM) and load it into a numpy array
> * place a light source at some coordinate
> * check for each point on the DEM if they can see the light source or if
they are being shadowed by other points of the DEM?
>
> As some kind of lightning calculations are done in the beautiful sceneries
of VPython, I wondered if I can 'abuse' it this way somehow?
>
> In case it can be done, I would need to know though, how accurate this
would be, as I would like to calculate solar irradiation times on the DEM
coordinates (just to make you interested! ;)
>
> Any suggestions welcome!

About 5 years ago I used `microdem` to estimate radio coverage of pre-wimax
equipment ... maybe its useful in your case.

C Anthony [mobile]

Re: [Visualpython-users] Raytracing VPython

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

But I don't think POVray will provide quantitative information about
the illumination at various places on the height field, will it?

Bruce Sherwood

On Wed, Mar 23, 2011 at 3:43 PM, Guy K. Kloss <guy...@au...> wrote:
> On Thu, 24 Mar 2011 04:07:13 Bruce Sherwood wrote:
>> I don't think VPython will help you at all. There is no shadowing in
>> VPython, that being difficult to do in real time. And currently there
>> is no distance dependence of the light from local lights (or from
>> distant lights).
>
> Maybe an export of a VPython scene with only one light source to POVray may do
> the trick? Just a thought ...
>
> 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...
>
> ------------------------------------------------------------------------------
> Enable your software for Intel(R) Active Management Technology to meet the
> growing manageability and security demands of your customers. Businesses
> are taking advantage of Intel(R) vPro (TM) technology - will your software
> be a part of the solution? Download the Intel(R) Manageability Checker
> today! http://p.sf.net/sfu/intel-dev2devmar
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users
>
>

Re: [Visualpython-users] Raytracing VPython

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

On Thu, 24 Mar 2011 04:07:13 Bruce Sherwood wrote:
> I don't think VPython will help you at all. There is no shadowing in
> VPython, that being difficult to do in real time. And currently there
> is no distance dependence of the light from local lights (or from
> distant lights).

Maybe an export of a VPython scene with only one light source to POVray may do 
the trick? Just a thought ...

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] Raytracing VPython

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

I don't think VPython will help you at all. There is no shadowing in
VPython, that being difficult to do in real time. And currently there
is no distance dependence of the light from local lights (or from
distant lights).

Bruce Sherwood

On Tue, Mar 22, 2011 at 11:47 PM, K. -Michael Aye
<mic...@sp...> wrote:
> Dear VPython folks,
>
> I was wondering if one can use VPython to do the following:
>
> * take a digital elevation map (DEM) and load it into a numpy array
> * place a light source at some coordinate
> * check for each point on the DEM if they can see the light source or if they are being shadowed by other points of the DEM?
>
> As some kind of lightning calculations are done in the beautiful sceneries of VPython, I wondered if I can 'abuse' it this way somehow?
>
> In case it can be done, I would need to know though, how accurate this would be, as I would like to calculate solar irradiation times on the DEM coordinates (just to make you interested! ;)
>
> Any suggestions welcome!
>
> Best regards,
> Michael
>
> ------------------------------------------------------------------------------
> Enable your software for Intel(R) Active Management Technology to meet the
> growing manageability and security demands of your customers. Businesses
> are taking advantage of Intel(R) vPro (TM) technology - will your software
> be a part of the solution? Download the Intel(R) Manageability Checker
> today! http://p.sf.net/sfu/intel-dev2devmar
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users
>

[Visualpython-users] Raytracing VPython

[K. -M. A. <mic...@sp...>]

Dear VPython folks,

I was wondering if one can use VPython to do the following:

* take a digital elevation map (DEM) and load it into a numpy array
* place a light source at some coordinate
* check for each point on the DEM if they can see the light source or if they are being shadowed by other points of the DEM?

As some kind of lightning calculations are done in the beautiful sceneries of VPython, I wondered if I can 'abuse' it this way somehow?

In case it can be done, I would need to know though, how accurate this would be, as I would like to calculate solar irradiation times on the DEM coordinates (just to make you interested! ;)

Any suggestions welcome!

Best regards,
Michael

[Visualpython-users] Python + OpenGL (Pycon talk 2011)

[kirby u. <kir...@gm...>]

Not directly Visual Python, but of potential interest to some here:

http://pycon.blip.tv/file/4882916/

Gives some pedagogical pointers, emphasizing virtues of
minimalist code, building a simple API.

Kirby

Some VPython resources:
http://www.4dsolutions.net/ocn/cp4e.html

Re: [Visualpython-users] Fedora

[Thomas S. <to...@fe...>]

On Mon, 21 Mar 2011 23:44:09 +0000
Peter Borcherds wrote:

> when visualpython-users are sending a message relevant only to a
> subject specific subgroup such as fedora, it would save the rest of
> us time if this were to be shown clearly in the message subject
> 
> Peter
> P H Borcherds
> ________________________________
> From: Thomas Spura [to...@fe...]
> Sent: 21 March 2011 22:11
> To: Bru...@nc...
> Cc: vis...@li...
> Subject: Re: [Visualpython-users] polygon dependency
> 
> On Mon, 21 Mar 2011 15:46:53 -0600
> Bruce Sherwood wrote:
> 
> > My mistake, as I had intended that the recent releases in fact make
> > Polygon and ttfquery optional (in which case extrusion and text
> > objects are not available). The mistake is that the start of
> > vis/shapes.py should look like this:
> >
> > from math import pi, cos, sin, sqrt, tan, atan, acos
> > try:
> >     from Polygon import Polygon
> >     from Polygon.Shapes import Star
> >     from ttfquery import describe, glyphquery, glyph
> > except:
> >     pass
> > import vis
> >
> > import os, sys, glob
> >
> > (vis/primitives.py already checks for Polygon and ttfquery
> > availability and issues an appropriate warning.)
> >
> > The corrected shapes.py is now in CVS.
> 
> 
> Thanks a lot... Then the next version will be usable in fedora
> again. :)
> 
>         Thomas

The polygon dependency is nasty on all free (as in freedom) operation
systems. A subject like "freeos: polygon dependency" doesn't make any
sense.

	Thomas

[Visualpython-users] Fedora

[Peter B. <p.h...@bh...>]

when visualpython-users are sending a message relevant only to a subject specific subgroup such as fedora, it would save the rest of us time if this were to be shown clearly in the message subject

Peter
P H Borcherds
________________________________
From: Thomas Spura [to...@fe...]
Sent: 21 March 2011 22:11
To: Bru...@nc...
Cc: vis...@li...
Subject: Re: [Visualpython-users] polygon dependency

On Mon, 21 Mar 2011 15:46:53 -0600
Bruce Sherwood wrote:

> My mistake, as I had intended that the recent releases in fact make
> Polygon and ttfquery optional (in which case extrusion and text
> objects are not available). The mistake is that the start of
> vis/shapes.py should look like this:
>
> from math import pi, cos, sin, sqrt, tan, atan, acos
> try:
>     from Polygon import Polygon
>     from Polygon.Shapes import Star
>     from ttfquery import describe, glyphquery, glyph
> except:
>     pass
> import vis
>
> import os, sys, glob
>
> (vis/primitives.py already checks for Polygon and ttfquery
> availability and issues an appropriate warning.)
>
> The corrected shapes.py is now in CVS.


Thanks a lot... Then the next version will be usable in fedora again. :)

        Thomas

------------------------------------------------------------------------------
Enable your software for Intel(R) Active Management Technology to meet the
growing manageability and security demands of your customers. Businesses
are taking advantage of Intel(R) vPro (TM) technology - will your software
be a part of the solution? Download the Intel(R) Manageability Checker
today! http://p.sf.net/sfu/intel-dev2devmar
_______________________________________________
Visualpython-users mailing list
Vis...@li...
https://lists.sourceforge.net/lists/listinfo/visualpython-users

Re: [Visualpython-users] polygon dependency

[Thomas S. <to...@fe...>]

On Mon, 21 Mar 2011 15:46:53 -0600
Bruce Sherwood wrote:

> My mistake, as I had intended that the recent releases in fact make
> Polygon and ttfquery optional (in which case extrusion and text
> objects are not available). The mistake is that the start of
> vis/shapes.py should look like this:
> 
> from math import pi, cos, sin, sqrt, tan, atan, acos
> try:
>     from Polygon import Polygon
>     from Polygon.Shapes import Star
>     from ttfquery import describe, glyphquery, glyph
> except:
>     pass
> import vis
> 
> import os, sys, glob
> 
> (vis/primitives.py already checks for Polygon and ttfquery
> availability and issues an appropriate warning.)
> 
> The corrected shapes.py is now in CVS.


Thanks a lot... Then the next version will be usable in fedora again. :)

	Thomas

Re: [Visualpython-users] polygon dependency

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

Installation on Linux is easy if there is a package in the package
manager, as is the case for example on Ubuntu. One problem is that at
least on Ubuntu (the only flavor of Linux with which I have
significant experience) the python-visual package tends to be quite
far out of date, though that hardly matters for a beginners' course.
Another problem is that apparently for many Linux distributions no one
has created an official package, and if you have to build from source
you need strong computer skills.

Bruce Sherwood

On Mon, Mar 21, 2011 at 3:38 PM, Seth Morton <sm...@ps...> wrote:
> Good luck!  I really enjoy vpython's simplicity. In fact, we are going to use it this summer to teach middle schoolers the basics of programming and science in a way that gives immediate feedback. The biggest problem I have is installation on linux (it's a breeze on Mac), though that's probably only a small population of the users, so I understand the slower development.
>
> Seth Morton

Re: [Visualpython-users] polygon dependency

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

My mistake, as I had intended that the recent releases in fact make
Polygon and ttfquery optional (in which case extrusion and text
objects are not available). The mistake is that the start of
vis/shapes.py should look like this:

from math import pi, cos, sin, sqrt, tan, atan, acos
try:
    from Polygon import Polygon
    from Polygon.Shapes import Star
    from ttfquery import describe, glyphquery, glyph
except:
    pass
import vis

import os, sys, glob

(vis/primitives.py already checks for Polygon and ttfquery
availability and issues an appropriate warning.)

The corrected shapes.py is now in CVS.

You've probably seen this, but on the download pages at vpython.org
you'll find this:

In support of the 3D text and extrusion objects, you will need to
install the font-handling modules FontTools, ttfquery (version 1.0.4
or later), and Polygon (all available from pypi.python.org), for which
the following conditions apply: "Polygon is being distributed subject
to the following conditions: This distribution contains code from the
GPC Library, and/or code resulting from the use of the GPC Library.
This usage has been authorized by The University of Manchester, on the
understanding that the GPC-related features are used only in the
context of this distribution. It is not permitted to extract the GPC
code from the distribution as the basis for commercial exploitation,
unless a GPC Commercial Use Licence is obtained from The University of
Manchester, contact: http://www.cs.man.ac.uk/~toby/gpc/".

Bruce Sherwood

On Mon, Mar 21, 2011 at 2:58 PM, Thomas Spura <to...@fe...> wrote:
> Hi list,
>
> is it possible to drop the polygon dependency somehow (or at least make
> it optionally).
>
> I just tried to package it for fedora and noticed the dependency to
> gpc, which is non-free:
> https://fedorahosted.org/fpc/ticket/74
>
> When that can't be resolved by polygon upstream (and I bet it can't,
> because they would be forced to rewrite the gpc library somehow),
> vpython will never work on fedora, because polygon can't be imported...
>
> Greetings,
>   Thomas
>
> ------------------------------------------------------------------------------
> Enable your software for Intel(R) Active Management Technology to meet the
> growing manageability and security demands of your customers. Businesses
> are taking advantage of Intel(R) vPro (TM) technology - will your software
> be a part of the solution? Download the Intel(R) Manageability Checker
> today! http://p.sf.net/sfu/intel-dev2devmar
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users
>

Re: [Visualpython-users] polygon dependency

[Seth M. <sm...@ps...>]

I agree. Typically, whenever an RPM is installed with Yum, it automatically installs all necessary dependencies, or at least enough so that the program will run. When I install python-visual after a clean install, I need to manually install gtkglextmm-devel, libglademm24-devel, and fonttools through yum, then Polygon and TTFQuery through easy_install. At this point I can run vpython with no import errors.  Even then, if I want 3D fonts, I have to find and package the Microsoft fonts from source or I get the error "there's something wrong with the font times". 

It would be nice if the RPM listed the development libraries required as dependencies to that yum could install them automatically. Additionally, if there could be some sort of try:  except ImportError: around the Polygon and TTFQuery imports, or only attempt to import them when the 3D fonts are requested, with some sort of error message like "Polygon and TTFQuery modules are required for 3D rendering of text" of they are not found. 

The problem is, as Thomas points out, that Fedora requires that all packages be completely free and open source, so unless all dependencies meet that criteria vpython won't run as currently implemented, but with a few tweaks (the try statement I mentioned earlier) it could run without problem unless 3D fonts are required. A possibility might be to add a background routine to install Polygon and TTFQuery on first run with 3D fonts in a way that is hidden from the user, similar to what Sage-math does. 

Good luck!  I really enjoy vpython's simplicity. In fact, we are going to use it this summer to teach middle schoolers the basics of programming and science in a way that gives immediate feedback. The biggest problem I have is installation on linux (it's a breeze on Mac), though that's probably only a small population of the users, so I understand the slower development. 

Seth Morton

On Mar 21, 2011, at 4:58 PM, Thomas Spura <to...@fe...> wrote:

> Hi list,
> 
> is it possible to drop the polygon dependency somehow (or at least make
> it optionally).
> 
> I just tried to package it for fedora and noticed the dependency to
> gpc, which is non-free:
> https://fedorahosted.org/fpc/ticket/74
> 
> When that can't be resolved by polygon upstream (and I bet it can't,
> because they would be forced to rewrite the gpc library somehow),
> vpython will never work on fedora, because polygon can't be imported...
> 
> Greetings,
>   Thomas
> 
> ------------------------------------------------------------------------------
> Enable your software for Intel(R) Active Management Technology to meet the
> growing manageability and security demands of your customers. Businesses
> are taking advantage of Intel(R) vPro (TM) technology - will your software 
> be a part of the solution? Download the Intel(R) Manageability Checker 
> today! http://p.sf.net/sfu/intel-dev2devmar
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users