visualpython-users

Re: [Visualpython-users] VPython Hangs

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

On Thu, 2005-03-17 at 21:32 -0300, Flavio Coelho wrote:
> I have an app which calls vpython to draw a graph. It runs perfectly
> on my gentoo box. But On my debian box, (Knoppix)

Is Knoppix running from CD-ROM, or from a hdd install?

>  it hangs silently,
> when it was supposed to   begin drawing my objects. I have been able
> to trace the failure (using Eric's debugger) to line 58 of
> primitives.py. Since the bug is apparently in the C++ code. I couldn't
> trace it further. The weirdest thing is that all demos run fine on th
> debian box...

Did you mean that the demos run fine on the Gentoo box, or the demos
work fine on the Debian machine and your program fails?

> As far as I could tell the versions of the main packages are the same
> on both boxes. The debian box doesn't have hardware GL acceleration
> though.

On the Debian system, does glxgears work (its in the xbase-clients
package)?

-Jonathan

[Visualpython-users] VPython Hangs

[Flavio C. <fcc...@gm...>]

I have an app which calls vpython to draw a graph. It runs perfectly
on my gentoo box. But On my debian box, (Knoppix) it hangs silently,
when it was supposed to   begin drawing my objects. I have been able
to trace the failure (using Eric's debugger) to line 58 of
primitives.py. Since the bug is apparently in the C++ code. I couldn't
trace it further. The weirdest thing is that all demos run fine on th
debian box...

As far as I could tell the versions of the main packages are the same
on both boxes. The debian box doesn't have hardware GL acceleration
though.


any hints?


--=20
Fl=E1vio Code=E7o Coelho
---------------------------
Great spirits have always found violent opposition from mediocrities. The
latter cannot understand it when a man does not thoughtlessly submit to
hereditary prejudices but honestly and courageously uses his intelligence.
        Albert Einstein

[Visualpython-users] Fedora Core 2 build problem

[Douglas S. B. <db...@br...>]

Hello,

New to VPython, searched the archives, but didn't see anything like 
this. Trying to build visual-3.0.3 from sources. The error is:

/usr/include/boost/python/type_id.hpp:69 error: syntax error before ')' 
token

Looks like BOOST_EXPLICIT_TEMPLATE_TYPE might not be defined?

Versions:

python 2.3.3
boost 1.31.0-7
linux 2.6.10-1.12_FC2

Any ideas would be appreciated,

-Doug

[Visualpython-users] Re: vpython

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

On Mon, 2005-03-14 at 13:12 -0500, Bruce Sherwood wrote:
> Jos=E9 A Mart=EDn H wrote:
> > Hi, is there any possibility to include an attribute to the curve objec=
t=20
> > that inform about its length
> > nor the quantity of points but the true length of the curve ?.
> >=20

See examples/drape.py:65-66, which almost does what you want.  Numeric
has the functionality you need without extending visual.curve directly.

> > it is easy to program it, but it could be very more intresting having=20
> > this feature as a native feature.

Why?  You can do it in 2-3 lines of code, and it won't be any slower
than what would go into a native Visual implementation.

> > =20
> > thanks
> > =20
> > Jose antonio Martin.

--=20
Jonathan Brandmeyer <jbr...@ea...>

Re: [Visualpython-users] deleting objects

[Hans F. <H.F...@so...>]

Dear Jonathan,

>> I was trying to delete visual python objects. Following the suggestions
>> made in this email
>> http://sourceforge.net/mailarchive/forum.php?thread_id=6392639&forum_id=3591,
>>
>> I thought this should be straightforward as follows
>>
>>
>> from visual import *
>>
>> x = sphere()
>>
>> #window appears, sphere appears in window
>>
>> del x    #delete x, therefore removing reference to x
>>
>> #here I expected the sphere to disappear but this was not the case.
>>
>> What am I getting wrong here?
>
> It appears you missed the point of the message you are referring to.
> Visual still owns a reference to the object, which you must destroy
> first by setting x.visible = False before deleting it.

I did indeed ;)

Thank you for clearing this up.

Hans

>
> -Jonathan
>
>
>
> -------------------------------------------------------
> SF email is sponsored by - The IT Product Guide
> Read honest & candid reviews on hundreds of IT Products from real users.
> Discover which products truly live up to the hype. Start reading now.
> http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users
>

Re: [Visualpython-users] deleting objects

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

On Mon, 2005-03-07 at 12:15 +0000, Hans Fangohr wrote:
> Dear all,
> 
> I was trying to delete visual python objects. Following the suggestions 
> made in this email 
> http://sourceforge.net/mailarchive/forum.php?thread_id=6392639&forum_id=3591,
> 
> I thought this should be straightforward as follows
> 
> 
> from visual import *
> 
> x = sphere()
> 
> #window appears, sphere appears in window
> 
> del x    #delete x, therefore removing reference to x
> 
> #here I expected the sphere to disappear but this was not the case.
> 
> What am I getting wrong here?

It appears you missed the point of the message you are referring to.
Visual still owns a reference to the object, which you must destroy
first by setting x.visible = False before deleting it.

-Jonathan

[Visualpython-users] deleting objects

[Hans F. <H.F...@so...>]

Dear all,

I was trying to delete visual python objects. Following the suggestions 
made in this email 
http://sourceforge.net/mailarchive/forum.php?thread_id=6392639&forum_id=3591,

I thought this should be straightforward as follows


from visual import *

x = sphere()

#window appears, sphere appears in window

del x    #delete x, therefore removing reference to x

#here I expected the sphere to disappear but this was not the case.

What am I getting wrong here?

Thank you,

Hans

Re: [Visualpython-users] stereo

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

On Sat, 2005-02-26 at 06:41 -0600, John Brawley wrote:
> Hi, Jonathan et.al.
> 
> Noting release of new VPython, wondering if/when another release will
> contain a modification to the stereo='passive' function, to make it
> selectable between "walleyed" stereo (as it now is) and "crosseyed" stereo.

I only heard about this feature request second-hand from Bruce a couple
of days ago.  I wanted to push out a new release before adding this
feature since several bugfixes have been added to CVS since the last
release, and I wanted to get them out to the public.  Adding this new
feature is on my todo list.

If you want to contribute a patch, the relevant files are
src/gldevice.cpp, src/display.cpp, include/gldevice.h and
include/display.h.

-Jonathan

Re: [Visualpython-users] Released: VPython 3.1.1

[Floris B. <fb...@so...>]

Hi

After the interuption of release 3.1.0 I've made the Debian packages
again for this release.  They're still for grabs from
http://www.soton.ac.uk/~fb102/Debian/ like before.

Please let me know any problems, (packaging) bugs or any other sort of
corrections and wishes on your mind.  And as before, SPARC binary
builds on demand.

Cheers
Floris

-- 
Debian GNU/Linux -- The power of freedom
www.debian.org | www.gnu.org | www.kernel.org

[Visualpython-users] stereo

[John B. <jb...@te...>]

Hi, Jonathan et.al.

Noting release of new VPython, wondering if/when another release will
contain a modification to the stereo='passive' function, to make it
selectable between "walleyed" stereo (as it now is) and "crosseyed" stereo.

I understand, from converse with Bruce (Sherwood) that the 'passive' stereo
function was written in walleyed because that was the _de-facto_ (not
'agreed-upon') 'standard' in the industry for _projected_ stereo, and that
amazingly enough, no thought _at_all_ was given to the community of stereo
"freeviewers."

There is an enormous amount of people out here who do "freeviewing"
(walleyed and/or crosseyed) stereo, and perhaps even many who, like me, do
it every day, and who _work_ in it.

Every freeviewing stereo computer modelmaking program I have ever seen,
_at_least_ offers the option to do the freeviewing either way.  Those
include Struck, Fluidiom, and Graph3D (to mention just the ones _I_ use
often), and a half-dozen others I have used but found wanting for one reason
or another.

I won't repeat all the arguments I made to Bruce, regarding why crosseyed
stereo is superior in many ways to walleyed for people who actually use it
in their studies, but I will mention that I think it was a piece of poorly
thought-out decisionmaking, to _exclude_ crosseyed-freeviewers from using
Visual Python's otherwise wonderfully simple-to-code stereo methodologies.

Bruce also mentioned he thought that making the option selectable between
walleyed and crosseyed would be a "minor" tweak to the code (it simply means
offering a way to switch the positions of the two images of the object, in
the frame), so I hope that at least by the next version release, you might
have seriously considered this, and made the option available for
crosseyed-freeviewers.

It would make no difference to those already using the 'passive' option for
projected (polarizer-using) stereo, and would open up VPython's stereo
options to those of us 'newbie' coders --in scientific studies or not-- who
know better than to try to use walleyed viewing on large monitors.  (It
would also not have impacted the 'standard' users of projected stereo, to
have made the stereo crosseyed in the _first_ place, since the stereo depth
is dependent solely on which eye one puts which polarizer in front of....)

Please?  Please?
Thanks!

Peace
JB
Xj...@te...
( REMOVE the 'X' before using ! )
NOTE: due to virus explosion, currently deleting ALL messages over 140kb
from the server without downloading.
http://tetrahedraverse.com

Re: [Visualpython-users] transparency

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

On Sat, 2005-02-26 at 04:45 -0300, Flavio Coelho wrote:
> Hi,
> 
> Is there a way to control the alpha channel (transparency) for objects
> in vpython?

Not in any released version.  If you want to try something experimental,
check out the "vpython-core2" module from CVS.  Most of the objects in
that module support transparency by setting color to an rgba 4-tuple, or
by setting object.alpha to some value.  However, there are some
significant parts missing from vpython-core2, particularly scene.mouse.

-- 
Jonathan Brandmeyer <jbr...@ea...>

[Visualpython-users] transparency

[Flavio C. <fcc...@gm...>]

Hi,

Is there a way to control the alpha channel (transparency) for objects
in vpython?

--=20
Fl=E1vio Code=E7o Coelho
---------------------------
Great spirits have always found violent opposition from mediocrities. The
latter cannot understand it when a man does not thoughtlessly submit to
hereditary prejudices but honestly and courageously uses his intelligence.
        Albert Einstein

[Visualpython-users] Released: VPython 3.1.1

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

VPython version 3.1.1 is available for download from our Sourceforge
project file release page.  Source tarballs for all platforms and
prebuilt packages for Windows are available.

If you particularly dislike the fact that there is no longer a
restriction for rotating upside-down, let us know.

The changes for this release follow.  I have also included the list of
changes for the 3.1.0 release, since I don't think that they were posted
to the mailing list at that time.

-Jonathan Brandmeyer

Visual 3.1.1
===============================================================================
    NEW FEATURES:
    * There is no longer a restriction on rotating past the up axis, 
        such that the scene may be rotated upside-down under user 
        control.
    * The MS Windows builds include packages for Python 2.3 and 2.4, 
        upgrades Boost.Python to version 1.32, and adds Numarray 1.1.1.
    * The MS Windows installer has been componentized, and will respect 
        an existing installation of Numeric or Numarray.	

    BUGS FIXED:
    * A resource leak in the label object has been fixed.
    * configure option --with-gtkgl-prefix works correctly (only 
        affects *nix and OSX users).
    * The bounds check for display.up and display.forward is more 
        robust.
    * A crash in the free function rotate( vector, angle, axis) has 
        been fixed.
    * Thin curve objects are rendered with the correct color in 
        anaglyph stereo rendering modes.
    * The "vpython" script is working again.
	
    KNOWN ISSUES:
    * "make install" is broken on Windows.  Use the generated
	MakeVPython.iss Inno Setup file to build an installer and run 
        it.

Visual 3.1.0
===============================================================================
    NEW FEATURES:
    * This is the first release that has been produced from our new CVS 
        layout.  This enables a more deterministic release process that 
        will reduce regressions.
    * Numarray support has been added.  You can enable support for 
        Numeric and Numarray in the same build, and select between them 
        at runtime.  See INSTALL.txt for details.

    BUGS FIXED:
    * The autotools support for MS Windows has been significantly 
        improved.
    * Nx2 arrays are supported for convex.pos and faces.pos
    * faces.color is initialized in the correct order.  This bug 
        prevented you from initializing faces.color to an array in the 
        constructor.
    * {primitive,frame}.rotate( pos=foo, origin=bar) works as expected.
    * The shutdown procedure is kinder to the shell when using Bash.
    * Several uninitialized value errors were fixed in the display 
        class when using the GTK+ backend.

Re: [Visualpython-users] using angle brackets for vectors

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

On Fri, 2005-02-18 at 08:44 -0600, Isaac W Hanson wrote:
> < 1e11, 0, 0 > isn't valid Python syntax by itself (although somewhat perversely, "( vector < 1e11, 0, 0 > vector )" is).
> 
> 
> Your last statement confuses me.  Isn't  "( vector < 1e11, 0, 0 > vector 
> )"  a tuple containing an expression (vector is-less-than 1e11), an 
> integer and another expression (zero is-greater-than vector)?  Perhaps I 
> missed something?
> 
>  - Isaac

Yes, that's right.  The __lt__ calls are evaluated right away, so the
result of the expression is a tuple "(False, 0, False)".  I was only
pointing out that it was syntacticly valid, not necessarily useful.

-Jonathan

Re: [Visualpython-users] using angle brackets for vectors

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

On Thu, 2005-02-17 at 15:06 -0500, Joe Heafner wrote:
> I'm running VPython 3.0 with Python 2.3.4 under Mac OS X 10.3.x. I just 
> noticed that this version of VPython uses angle brackets when printing 
> out the components of vectors, e.g. <1,2,3>. Previous versions used 
> square brackets. Since the M&I textbook uses angle brackets, I was 
> wondering whether it would be possible, 

To the best of my knowledge, the use of parenthesis in Python in this
way cannot be changed.  I changed the way that vectors are converted to
strings for the reasons you list below, but I'm pretty sure that there
isn't a way to introduce new syntactic elements to Python in this way.

> or even worthwhile, to change 
> the VPython syntax to use angle brackets to define the numerical 
> components of vectors. For example, instead of writing cart.pos = 
> vector(2,0,0)
> we could write cart.pos = <2,0,0> or instead of writing Earth = 
> sphere(pos=vector(1e11,0,0)) we could write Earth = 
> sphere(pos=<1e11,0,0>). Basically any time the components of a vector 
> quantity are given as a triplet of numbers the angle brackets could be 
> used.

> Pedagogically, I think this would make VPython syntax "look" more like 
> the notation in the textbook. While the "vector" preceding the 
> components in the pos attribute above is optional, 

It is optional only in the sense that any two- or three-sequence of
numbers is implicitly converted to a vector when calling any function in
visual itself.  The statement sphere( pos=(1e11, 0, 0)) creates a
temporary tuple (1e11, 0, 0) and passes it to the sphere object's
constructor, which is automatically changed to a vector by Visual.  For
similar reasons, the statement sphere( pos=[1e11, 0, 0]) works.  < 1e11,
0, 0 > isn't valid Python syntax by itself (although somewhat
perversely, "( vector < 1e11, 0, 0 > vector )" is).

> I tell my students 
> to include it. I think using angle brackets would reduce the list of 
> things to remember by one and still immediately draw attention to 
> vectors in VPython code.

-Jonathan

[Visualpython-users] using angle brackets for vectors

[Joe H. <hea...@ct...>]

I'm running VPython 3.0 with Python 2.3.4 under Mac OS X 10.3.x. I just 
noticed that this version of VPython uses angle brackets when printing 
out the components of vectors, e.g. <1,2,3>. Previous versions used 
square brackets. Since the M&I textbook uses angle brackets, I was 
wondering whether it would be possible, or even worthwhile, to change 
the VPython syntax to use angle brackets to define the numerical 
components of vectors. For example, instead of writing cart.pos = 
vector(2,0,0)
we could write cart.pos = <2,0,0> or instead of writing Earth = 
sphere(pos=vector(1e11,0,0)) we could write Earth = 
sphere(pos=<1e11,0,0>). Basically any time the components of a vector 
quantity are given as a triplet of numbers the angle brackets could be 
used.

Pedagogically, I think this would make VPython syntax "look" more like 
the notation in the textbook. While the "vector" preceding the 
components in the pos attribute above is optional, I tell my students 
to include it. I think using angle brackets would reduce the list of 
things to remember by one and still immediately draw attention to 
vectors in VPython code.

Cheers,
Joe Heafner -- Astronomy/Physics Instructor (by some definitions)

[Visualpython-users] closing vpython scene

[<ku...@ae...>]

I have upgraded to Python 2.4 and latest vpython 3.1.0 for 2.4 from python
2.3.4
My codes seemed to run well without any inconsistency unfortunately I
realized that when I closed the vpython scenes from the window X buttons,
they terminated the entire Tix application which called them. I am sure the
scene.exit=0 is not functioning well in this matter but could not find any
solutions.
I already tried the same thing on another computer but the problem persists.

Can anyone help me close the vpython scene without closing Python gui
application which called
it?

[Visualpython-users] scene.mouse problem in experimental release?

[Jacob S. <ke...@ja...>]

Hi I'm new here, not to python.

You're looking for bugs in the experimental release, right?
Running this code

=====================
from visual import *
box()
while 1:
    if scene.mouse.button == 'left':
        break
scene.visible = 0
=======================

yields the error

============================
Traceback (most recent call last):
    File "<stdin>", line 2, in ?
ValueError: Button type should be left, right, or middle
============================

This happens whenever I try to access *any* scene.mouse attributes

HTH,
Jacob Schmidt

Re: [Visualpython-users] vpython install on Slackware10, numarray instead of Numeric

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

On Wed, 2005-01-26 at 17:16 -0800, Eric Ayars wrote:

> There are still a couple configuration problems to clean up, though. 
> For some reason, the vpython startup shell script doesn't know where 
> things are, yet. I edited it to hard-code the directories:
> 	edit /usr/bin/vpython
> 		change "@visualdemodir@" to 
> "/usr/lib/python2.3/site-packages/visual/examples"
> 		change "@visualidledir@" to "/usr/bin"
> 		change "@IDLE@" to "idle"
> There is undoubtedly a better way to go about this. Probably I needed 
> to set something on the configure step of the vpython installation, but 
> I don't know what and this worked. (Suggestions?)

How about: "Tell Jonathan that its broken."  You shouldn't need to do
this: configure is supposed to do it for you.  It will be fixed soon.

-Jonathan

[Visualpython-users] vpython install on Slackware10, numarray instead of Numeric

[Eric A. <Ay...@ma...>]

I've installed visual 3.1 on my Slackware 10.0 boxes, and gotten it to 
work! Here are some tricks I had to use, just in case someone else is 
running into problems doing the same on a stock Slackware x86 install.

Please note: this is what I DID. I'm sure there may be a better way to 
do it, as I'm still a relative newbie to linux; but this worked.

After a "install everything" Slackware install (from CD):
	install gtkglarea
		download gtkglarea-1.2.2.tar.gz from Twocows.
		$ tar -xzf gtkglarea-1.2.2.tar.gz
		$ cd gtkglarea-1.2.2/
		$ ./configure
		$ make
		$ sudo make install
		$ make clean
	install numarray (successful installation of Numeric on Slackware has 
so far eluded me.)
		download numarray-1.1.1.tar.gz from SourceForge
		$ tar -xzf numarray-1.1.1.tar.gz
		$ cd numarray-1.1.1/
		$ sudo python setup.py install
	get boost-jam
		download boost-jam-3.1.10-1-linuxx86.tgz from SourceForge
		$ tar -xzf boost-jam...
	install boost libraries
		download boost_1_32_0.tar.gz from SourceForge
		$ tar -xzf boost_1_32...
		$ cd boost_1_32_0/
		$ sudo ../boost-jam.../bjam -sTOOLS=gcc -sPYTHON_ROOT=/usr \
			-sPYTHON_VERSION=2.3 install
		At this point, you may want to make a cup of coffee. If you have a 
slow machine, start with a coffee-plant seedling.
		$ cd /usr/local/lib/
		$ sudo ln -s libboost_python-gcc-mt-1_32.so 
/usr/local/lib/libboost_python.so
	install vpython
		Download visual-3.1.0.tar.gz (SourceForge again)
		$ tar -xzf visual-3.1.0.tar.gz
		$ cd visual-3.1.0/
		$ CPPFLAGS=-I/usr/local/include/boost-1_32
		$ export CPPFLAGS
		$ LDFLAGS=-L/usr/local/lib
		$ export LDFLAGS
		$ ./configure --prefix=/usr
		$ make
		$ sudo make install
		$ make clean

There are still a couple configuration problems to clean up, though. 
For some reason, the vpython startup shell script doesn't know where 
things are, yet. I edited it to hard-code the directories:
	edit /usr/bin/vpython
		change "@visualdemodir@" to 
"/usr/lib/python2.3/site-packages/visual/examples"
		change "@visualidledir@" to "/usr/bin"
		change "@IDLE@" to "idle"
There is undoubtedly a better way to go about this. Probably I needed 
to set something on the configure step of the vpython installation, but 
I don't know what and this worked. (Suggestions?)
		
There is just one more problem, which will show up if (like me) you've 
been unable to install Numeric.
The module primitives.py 
(/usr/lib/python2.3/site-packages/visual/primitives.py) still tries to 
import Numeric, even after the module(s) that import primitives.py go 
through a bit of trouble to determine whether you have Numeric or not. 
This should probably be fixed in a future revision: for now I got 
around it by changing "Numeric" to "numarray" in line 10 of 
primitives.py.

Thanks to Jonathan Brandmeyer for his help, and I hope this summary is 
of use to someone...

-EA

-----------------------------------------------------------------
Dr. Eric Ayars
Assistant Professor of Physics
California State University, Chico
ay...@ma...

[Visualpython-users] stereo.py

[John B. <jb...@te...>]

> Subject: [Visualpython-users] Inherent Stereo

> FYI, anyone trying to do real 3D stereo:
> There is a nifty simple "kludge" that gives stereo without additional
code.
[snips]

>Date: Sun, 23 Jan 2005 15:21:33 -0500
>From: Bruce Sherwood <Bru...@nc...>
>A particularly easy way to achieve this effect in an existing program is
>to add the statement
>
>scene.stereo = 'passive'

Please, Bruce: _just_ that statement?  Nothing else?

>For a specific example, see stereo.py in the "Contributed programs"
>section of vpython.org. And note that scene.stereodepth can be used to
>move the scene forward. Further details are in the section on
>"Controlling windows" in the on-line reference manual.

I've examined and tried to plug pieces of stereo.py into my program.
I get very odd results:
I can get scene1 (say, 'left') to work as the program works; balls appear,
and change positions just as they're supposed to, but
"copyobjects(left,right)" gives me a) an occasional color-changing HUGE ball
occupying most of the screen, and b) no erasure of previous balls copied
from the active scene ('left').

I also can't understand what's being done by the def:
copyobjects(left,right), and meddling with it seems only to make things
worse....

I'm stuck.  (Been here before: a year or more ago you recommended similar
methods to me, and I could not make them work then, either.)  Is there an
explanation ("for dummies") of what copyobjects(left,right) is actually
doing, anywhere that I can find it?

stereo.py runs perfectly by itself (the demo, demonstrates).

I have only one need for the kind of stereo you recommended here, (and I
await knowing if " scene.stereo = 'passive' " operates alone without
stereo.py), and that one thing is the ability to rotate and zoom on my
objects in both windows simultaneously.  Otherwise I'd stick with my
"kludge" because it's probably roughly 100 times faster, without hiccups and
pauses, than the stereo.py method.  (I simply draw the same objects twice,
once into 'left' and once into 'right', so every time a ball moves, it moves
in both scenes 'simultaneously', and since one of the scenes uses a
different center, the stereo has nothing to do with the drawing.)

Please: does scene.stereo='passive' work alone (no other statements), and
how can I find out what copyobjects(a,b) is doing?  (I have never had to use
any __double-underlined__ statements before....)

Thanks for the advice so far!

Peace
JB
Xj...@te...
( REMOVE the 'X' before using ! )
NOTE: due to virus explosion, currently deleting ALL messages over 140kb
from the server without downloading.
http://tetrahedraverse.com

[Visualpython-users] New contributed programs

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

In the user contributed programs area of vpython.org is a link to an 
extensive set of Earth Science programs written by Lensyl Urbano of the 
Dept. of Earth Sciences at the University of Memphis.

Bruce Sherwood

Re: [Visualpython-users] Inherent Stereo

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

A particularly easy way to achieve this effect in an existing program is 
to add the statement

scene.stereo = 'passive'

For a specific example, see stereo.py in the "Contributed programs" 
section of vpython.org. And note that scene.stereodepth can be used to 
move the scene forward. Further details are in the section on 
"Controlling windows" in the on-line reference manual.

Bruce Sherwood

John Brawley wrote:

> FYI, anyone trying to do real 3D stereo:
> There is a nifty simple "kludge" that gives stereo without additional code.
> 
> You must of course use two side-by-side windows, with the usual
> "scene=display(title..."(etc.) code.
> 
> First window's "center=" numbers should be (0.35,0,0)
> Second window's "center=" numbers: (0,0,0)
> 
> This will very slightly offset one window's contents from the other's, but
> it also gives real 3D depth to what's in the scene.  The 'camera' apparently
> shifts somewhat in the 'x' direction, which is enough to do the necessary
> brain-fooling trick. ( IOW, I don't know _why_ this works, only that it
> _does_ work.)
> 
> To increase or decrease the apparent depth ("hyperstereo" or less depth),
> slightly increase or decrease the first number ("0.35" in the above
> example).
> To invert the stereo (use "walleyed" freeviewing instead of crosseyed
> feeviewing), make that number negative.
> 
> Now, I can't guarantee that it'll work properly with just _every_
> conceivable object or set of objects, but it works perfectly with my
> particular object-set, which is a grouping of small spheres being compressed
> together inside an invisible larger sphere.
> 
> I discovered this when trying to learn to use VPython, and having enormous
> difficulty trying to find and use code that would give me stereo.  VPython's
> windowing system is probably not _designed_ to do this, but it does do it,
> and does it without a shred of "stereo-specific" additional code.
> 
> Just passing learned stuff to the list.
> Thank me only if you a) want to do stereo, and b) try it and it works for
> you as well.
> (*grin*)
> 
> John Brawley
> 
> 
> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting
> Tool for open source databases. Create drag-&-drop reports. Save time
> by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc.
> Download a FREE copy at http://www.intelliview.com/go/osdn_nl
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users

[Visualpython-users] Inherent Stereo

[John B. <jgb...@ch...>]

FYI, anyone trying to do real 3D stereo:
There is a nifty simple "kludge" that gives stereo without additional code.

You must of course use two side-by-side windows, with the usual
"scene=display(title..."(etc.) code.

First window's "center=" numbers should be (0.35,0,0)
Second window's "center=" numbers: (0,0,0)

This will very slightly offset one window's contents from the other's, but
it also gives real 3D depth to what's in the scene.  The 'camera' apparently
shifts somewhat in the 'x' direction, which is enough to do the necessary
brain-fooling trick. ( IOW, I don't know _why_ this works, only that it
_does_ work.)

To increase or decrease the apparent depth ("hyperstereo" or less depth),
slightly increase or decrease the first number ("0.35" in the above
example).
To invert the stereo (use "walleyed" freeviewing instead of crosseyed
feeviewing), make that number negative.

Now, I can't guarantee that it'll work properly with just _every_
conceivable object or set of objects, but it works perfectly with my
particular object-set, which is a grouping of small spheres being compressed
together inside an invisible larger sphere.

I discovered this when trying to learn to use VPython, and having enormous
difficulty trying to find and use code that would give me stereo.  VPython's
windowing system is probably not _designed_ to do this, but it does do it,
and does it without a shred of "stereo-specific" additional code.

Just passing learned stuff to the list.
Thank me only if you a) want to do stereo, and b) try it and it works for
you as well.
(*grin*)

John Brawley

Re: [Visualpython-users] destroying objects

[Flavio C. <fcc...@gm...>]

Hi Johnatan,

I followed your suggestion and no longer recreate  the ticklabels,
just update their "text" property. It is much faster now! Thanks!

The memory leak is still there but much smaller now. I look forward to
test my code again with the next release.

As usual, I posted the latest version of the stripchart module along
with a screenshot  here:

http://www.procc.fiocruz.br:8080/procc/Members/flavio/codepy/strip/file_vie=
w

If anyone on the list is interested in testing, please help
yourselves, and notify me of any bugs. feature requests are also
welcome.


> Found.  It was an internal bug in Visual's label and font deletion code.
> The fix will be in the next release.
>=20
> Actually, I'm a little surprised that this hasn't affected anyone else
> before this.  As far as I can tell, it has been in there since VPython's
> original release!
>=20
> Thank you for your useful reports and feedback.

I am glad I could help!

cheers,



--=20
Fl=E1vio Code=E7o Coelho
---------------------------
Great spirits have always found violent opposition from mediocrities. The
latter cannot understand it when a man does not thoughtlessly submit to
hereditary prejudices but honestly and courageously uses his intelligence.
        Albert Einstein