[PyOpenGL-Users] Re: Build notes...
Brought to you by:
mcfletch
Menu
▾
▴
[PyOpenGL-Users] Re: Build notes...
|
From: <il...@ya...> - 2002-07-20 06:22:09
|
Hi Mike,
this is in reply to your message about testing before.
. Updated with some notes about the new changes in
the cvs. Thought I'd cc it to the list in case there
is someone there who wants to help out :)
The current cvs needs swig 1.3.13 to build wrappers
and compile.
Now the pragma(python) code = "" has been replaced by
the new style shadow %{ %}, most of the problems seem
to be gone.
However a few problems remain, so I've gone through
the list of things you tested, and tested them again.
I wrote something like 'works' if it was working, else
a description of the problem, and a TODO note.
If no one wants to work on the problems below, I'll
start working on them next weekend :)
For the moment I'm working on some bugs and patches
from sourceforge, and the mailing list.
--- "Mike C. Fletcher" <mcf...@ro...> wrote: >
(Basically successful build, 1 major error (missing
> gluUnproject,
> gluProject, seem just to be missing %name directives
> in the wrapper),
TODO: On windows, see if this still happens? Works ok
here(debian/linux).
>a
> strange result with the array-based tests (point
> appears in wildly
> inappropriate position), and a few failures which
> may or may not be
> related to the conversion (I didn't normally run the
> OpenGL tests, Tarn
> did, so I'm not sure the code was correct before)).
>
> Okay, must get some other work done. Hope things
> are good with you,
> Mike
>
>
> Log of building and testing on Python 2.2.1 with
> VC++6 starts here...
>
> Uninstall PyOpenGL 2.0.0.45, delete package
> directory.
>
> Okay, have just installed swig 1.3.13, put swig.bat
> in a directory on
> the path, pointing to the installation directory and
> am proceeding to build.
>
> First problem, without Tk installed, setup.py bombs
> out. Probably
> should have a try/except around import of
> togl_setup.
>
> Okay, installed Python+Tkinter again from installer,
> killed old Tk
> environmental variables and installation, trying
> again.
>
> setup.py build_w
>
> Fails, saying that it can't find swig. Typing swig
> at the prompt starts
> swig, swig takes variables. Turns out the distutils
> script assumes that
> a programme on windows will use .exe, so it doesn't
> accept the .bat .
> Solution is to add swig install directory directly
> to path.
>
> setup.py build_w
>
> Now runs to completion.
>
> setup.py build
>
> Generates a considerable number of warnings for the
> GLE module, for
> instance:
>
> src/interface/GLE.0300.inc(1080) : warning C4024:
> 'gleExtrusion' :
> different types for formal and actual parameter 7
> src/interface/GLE.0300.inc(1156) : warning C4090:
> 'function' : different
> 'const' qualifiers
>
> (I consider this notable because of the failure of a
> few of the gle
> tests (notably the twburton one) during testing)
>
Not sure of these problems yet. Be good to see if it
is happening
in the old version.
I'm not sure how many people are interested in this
library?
Anyone interested in fixing it up? If not, I'd be
happy to leave
it as is, and note in the docs that it is
unmaintained.
> then eventually fails because tcl.h can't be found.
>
> tcl.h is not available from the Python installation,
> so go to
> ActiveState and download Active TCL 8.3.4.3
> installer. Add tcl include
> directory to environment's include setting.
>
I think we need some docs to describe how to build it.
- required libraries, how to set up system etc.
Do these exist allready?
TODO: write build docs, if not allready existing.
Also need to make the tk/tcl stuff optional.
TODO: make build path without tcl( this
possible/desirable? ).
> setup.py build
>
> No go, still can't find tcl.h, is apparently only
> looking in the
> explicitly specified directories, so I copy tcl.h
> and tk.h into the
> Python include directory.
>
> setup.py build
>
> No go, the files include other files from the tcl
> include directory, so
> I copy the whole lot of them into the Python include
> directory,
> including the X11 subdirectory.
>
> setup.py build
>
> Okay, that compiles all the way through.
>
> setup.py install
>
> Installation succeeds.
> Edit the __init__ file to not import the missing
> values as you suggested.
>
Not needed anymore. The __numeric_* stuff comes
through properly.
> Try to run some OpenGLContext tests.
> All fail, the gluUnproject function is missing (the
> Context uses this
> for the navigation mechanisms, so every Context
> sub-class fails to load).
Ran a number of these.
All seemed fine(except array_landscape.py)
Traceback (most recent call last):
File "array_landscape.py", line 30, in ?
base = landscape (2,2)
File "array_landscape.py", line 14, in landscape
base [:,:,1] = heights
ValueError: matrices are not aligned for copy
TODO: check/fix this error.
>
> Switch to running the non-Context tests...
> dots.py -- works
works.
> dek/* -- works, but missing gluUnproject, so can't
> interact
works now, can interact.
> gle/maintest -- silently exits
This one is a framework for the other tests.
> gle/texas -- builds window, but no visible geometry
TODO: see if gle/texas.py runs ok with last version.
see why not showing anything.
> gle/* -- others work
works.
>
>
C:\bin\lang\py22\Lib\site-packages\OpenGL\Demo\GLUT\examples>molehill.py
> Traceback (most recent call last):
> File
>
"C:\bin\lang\py22\Lib\site-packages\OpenGL\Demo\GLUT\examples\molehill.py
> ", line 153, in ?
> main()
> File
>
"C:\bin\lang\py22\Lib\site-packages\OpenGL\Demo\GLUT\examples\molehill.py
> ", line 122, in main
> gluNurbsSurface(nurb, knots, knots, pts1,
> GL_MAP2_VERTEX_3)
> OpenGL.GLU.GLUerror: [Errno 100284] negative
> byte-stride encounteed
TODO: look at this and fix. I still get same error.
>
> nehe/* -- all work as expected
works ok here.
> redbook/* -- all work as expected
works.
> simple/* -- all work as expected
works. this is a GLE test, looks like the helix one.
TODO: see if same as GLE/helix.py one. if so remove.
> tom/arraytest, poly -- weird spikes into the
> triangle/loop, not sure
> why, but looks like an array storage problem, or
> maybe a problem in the
> loop code
looks ok here. it's a multi colored circle. can you
give me a screen shot if it's still happening?
TODO: test on Mikes system.
> tom/demo -- seems to have tk menu problems (not
> likely OpenGL related)
>
works ok here.
>
C:\bin\lang\py22\Lib\site-packages\OpenGL\Demo\tom>fog.py
> Traceback (most recent call last):
> File
>
"C:\bin\lang\py22\Lib\site-packages\OpenGL\Demo\tom\fog.py",
>
> line 86, in
> ?
> main()
> File
>
"C:\bin\lang\py22\Lib\site-packages\OpenGL\Demo\tom\fog.py",
>
> line 84, in
> main
> f.run()
> File
>
"C:\bin\lang\py22\Lib\site-packages\OpenGL\Demo\tom\fog.py",
>
> line 32, in
> run
> self.init()
> File
>
"C:\bin\lang\py22\Lib\site-packages\OpenGL\Demo\tom\fog.py",
>
> line 59, in
> init
> glFogf(GL_FOG_COLOR, fogColor)
> TypeError: bad argument type for built-in operation
>
I get the same error.
TODO: look at this, try and fix. compare to old
version.
> All of the "tom" demos are without gluUnproject
> support, so no testing
> of left-click functionality is possible.
>
>
C:\bin\lang\py22\Lib\site-packages\OpenGL\Demo\twburton>knot.py
> Traceback (most recent call last):
> File
>
"C:\bin\lang\py22\Lib\site-packages\OpenGL\Demo\twburton\knot.py",
> line 1
> 96, in onDisplay
> glePolyCylinder(self.knot[i] +
> self.knot[i][:3], None,
> self.radius/15.0)
> ValueError: frames are not aligned
>
I still get the same error.
TODO: find someone who wants to maintain the GLE
library. else mark as unmaintained.
__________________________________________________
Do You Yahoo!?
Everything you'll ever need on one web page
from News and Sport to Email and Music Charts
http://uk.my.yahoo.com
|
